# Semantic conventions for generative AI inference token metrics **Status**: [Development][DocumentStatus] - [Semantic conventions for generative AI inference token metrics](#semantic-conventions-for-generative-ai-inference-token-metrics) - [Inference token metrics](#inference-token-metrics) - [Metric: `gen_ai.client.inference.usage.input_tokens`](#metric-gen_aiclientinferenceusageinput_tokens) - [Metric: `gen_ai.client.inference.usage.output_tokens`](#metric-gen_aiclientinferenceusageoutput_tokens) - [Metric: `gen_ai.client.inference.usage.cache_read.input_tokens`](#metric-gen_aiclientinferenceusagecache_readinput_tokens) - [Metric: `gen_ai.client.inference.usage.cache_write.input_tokens`](#metric-gen_aiclientinferenceusagecache_writeinput_tokens) - [Metric: `gen_ai.client.inference.usage.reasoning.output_tokens`](#metric-gen_aiclientinferenceusagereasoningoutput_tokens) - [Metric: `gen_ai.client.inference.operation.input_tokens`](#metric-gen_aiclientinferenceoperationinput_tokens) - [Metric: `gen_ai.client.inference.operation.output_tokens`](#metric-gen_aiclientinferenceoperationoutput_tokens) ## Inference token metrics The conventions described in this document are specific to Generative AI client applications and describe how inference token metrics are reported. Individual systems may include additional system-specific attributes. It is recommended to check system-specific documentation, if available. Two families of token instruments are defined: - **Usage** (`gen_ai.client.inference.usage.*` counters) are the primary instruments for measuring token consumption. They track how many tokens were spent over time, spending rate, and serve as a proxy for cost approximation. Each token category is a separate monotonic counter broken down by [`gen_ai.token.modality`](/docs/registry/attributes/gen-ai.md). The `cache_read`, `cache_write`, and `reasoning` counters report subsets of the `input_tokens` and `output_tokens` counters. - **Per-operation token distribution** (`gen_ai.client.inference.operation.input_tokens` and `gen_ai.client.inference.operation.output_tokens`) describe how token usage is distributed across individual inference operations. They should not be used for total usage or cost calculations. Use histograms exclusively to monitor usage percentiles (e.g. p95/p99) and detect outliers. Embeddings operations only report the [`gen_ai.client.embeddings.operation.input_tokens`](client-embeddings.md#metric-gen_aiclientembeddingsoperationinput_tokens) histogram, and total token usage can be computed from it. > [!NOTE] > > The `gen_ai.client.inference.operation.input_tokens` and > `gen_ai.client.inference.operation.output_tokens` histograms are > deliberately not broken down by modality: percentiles across modalities do > not add up. > > A query like this one appears valid: > > ```promql > histogram_quantile(0.95, > sum(rate({"gen_ai.client.inference.operation.input_tokens"}[5m])) > ) > ``` > > However, if modality were added as a dimension, it would produce meaningless results: for > 100 identical operations with 300 input tokens each (100 text and 200 image), > it would return 200 instead of 300. > > For the full reasoning behind this design, see > [Token metrics design](/docs/gen-ai/non-normative/token-metrics-design.md). ### Metric: `gen_ai.client.inference.usage.input_tokens` | Name | Instrument Type | Unit (UCUM) | Description | Stability | Entity Associations | | -------- | --------------- | ----------- | -------------- | --------- | ------ | | `gen_ai.client.inference.usage.input_tokens` | Counter | `{token}` | The number of input (prompt) tokens used, including cached tokens. [1] |  | | **[1]:** This metric SHOULD be reported when an operation involves the usage of tokens and the count is readily available. When systems report both used tokens and billable tokens, instrumentation MUST report billable tokens. Instrumentation SHOULD record contributions from successful and failed inference operations. Each measurement MUST be a known, strictly positive modality contribution. When a contribution's modality cannot reliably be determined, instrumentation MUST set `gen_ai.token.modality` to `unknown`. **Requirement level:** [Recommended](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/general/signal-requirement-level.md). **Attributes:** | Key | Stability | [Requirement Level](https://opentelemetry.io/docs/specs/semconv/general/attribute-requirement-level/) | Value Type | Description | Example Values | | --- | --- | --- | --- | --- | --- | | [`gen_ai.operation.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The name of the operation being performed. [1] | `chat`; `generate_content`; `text_completion` | | [`gen_ai.provider.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The Generative AI provider as identified by the client or server instrumentation. [2] | `openai`; `gcp.gen_ai`; `gcp.vertex_ai` | | [`gen_ai.token.modality`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The modality of the tokens being counted. [3] | `text`; `image`; `audio` | | [`gen_ai.request.model`](/docs/registry/attributes/gen-ai.md) |  | `Conditionally Required` If available. | string | The name of the GenAI model a request is being made to. [4] | `gpt-4` | | [`server.port`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Conditionally Required` If `server.address` is set. | int | GenAI server port. [5] | `80`; `8080`; `443` | | [`gen_ai.response.model`](/docs/registry/attributes/gen-ai.md) |  | `Recommended` | string | The name of the model that generated the response. | `gpt-4-0613` | | [`server.address`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Recommended` | string | GenAI server address. [6] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | **[1] `gen_ai.operation.name`:** If one of the predefined values applies, but specific system uses a different name it's RECOMMENDED to document it in the semantic conventions for specific GenAI system and use system-specific name in the instrumentation. If a different name is not documented, instrumentation libraries SHOULD use applicable predefined value. **[2] `gen_ai.provider.name`:** Semantic conventions for individual GenAI operations SHOULD clarify which kinds of providers (e.g. inference, embeddings, retrieval, memory, hosted agent providers) apply when it is not clear from context. The attribute SHOULD be set based on the instrumentation's best knowledge and may differ from the actual upstream provider. For example, a client SDK may be configured against a proxy or hosting platform that transparently relays requests to a different provider. The `gen_ai.provider.name` attribute acts as a discriminator that identifies the GenAI telemetry format flavor specific to that provider within GenAI semantic conventions. It SHOULD be set consistently with provider-specific attributes and signals. For example, GenAI spans, metrics, and events related to AWS Bedrock should have the `gen_ai.provider.name` set to `aws.bedrock` and include applicable `aws.bedrock.*` attributes and are not expected to include `openai.*` attributes. **[3] `gen_ai.token.modality`:** When the provider does not break usage down by modality and the modality cannot be reliably determined, instrumentations SHOULD report the tokens under the `unknown` modality. **[4] `gen_ai.request.model`:** The name of the GenAI model a request is being made to. If the model is supplied by a vendor, then the value must be the exact name of the model requested. If the model is a fine-tuned custom model, the value should have a more specific name than the base model that's been fine-tuned. **[5] `server.port`:** When observed from the client side, and when communicating through an intermediary, `server.port` SHOULD represent the server port behind any intermediaries, for example proxies, if it's available. **[6] `server.address`:** When observed from the client side, and when communicating through an intermediary, `server.address` SHOULD represent the server address behind any intermediaries, for example proxies, if it's available. --- `gen_ai.operation.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `chat` | Chat completion operation such as [OpenAI Chat API](https://platform.openai.com/docs/api-reference/chat) |  | | `create_agent` | Create GenAI agent |  | | `create_memory` | Create new memory records |  | | `create_memory_store` | Create or initialize a memory store |  | | `delete_memory` | Delete memory records |  | | `delete_memory_store` | Delete or deprovision a memory store |  | | `embeddings` | Embeddings operation such as [OpenAI Create embeddings API](https://platform.openai.com/docs/api-reference/embeddings/create) |  | | `execute_tool` | Execute a tool |  | | `fetch_response` | Fetch a previously generated model response by its identifier, without performing inference, such as [OpenAI Get a model response](https://platform.openai.com/docs/api-reference/responses/get) [7] |  | | `generate_content` | Multimodal content generation operation such as [Gemini Generate Content](https://ai.google.dev/api/generate-content) |  | | `invoke_agent` | Invoke GenAI agent |  | | `invoke_workflow` | Invoke GenAI workflow |  | | `plan` | Agent planning or task decomposition phase |  | | `retrieval` | Retrieval operation such as [OpenAI Search Vector Store API](https://platform.openai.com/docs/api-reference/vector-stores/search) |  | | `search_memory` | Search/query memories from a memory store |  | | `text_completion` | Text completions operation such as [OpenAI Completions API (Legacy)](https://platform.openai.com/docs/api-reference/completions) |  | | `update_memory` | Update existing memory records |  | | `upsert_memory` | Create or update memory records without the caller choosing which |  | **[7]:** Instrumentations SHOULD NOT report token usage (as attributes or metrics) for this operation. --- `gen_ai.provider.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `anthropic` | [Anthropic](https://www.anthropic.com/) |  | | `aws.bedrock` | [AWS Bedrock](https://aws.amazon.com/bedrock) |  | | `azure.ai.inference` | Azure AI Inference |  | | `azure.ai.openai` | [Azure OpenAI](https://learn.microsoft.com/en-us/azure/ai-services/openai/overview) |  | | `cohere` | [Cohere](https://cohere.com/) |  | | `deepseek` | [DeepSeek](https://www.deepseek.com/) |  | | `gcp.gemini` | [Gemini](https://cloud.google.com/products/gemini) [8] |  | | `gcp.gen_ai` | Any Google generative AI endpoint [9] |  | | `gcp.vertex_ai` | [Vertex AI](https://cloud.google.com/vertex-ai) [10] |  | | `groq` | [Groq](https://groq.com/) |  | | `ibm.watsonx.ai` | [IBM Watsonx AI](https://www.ibm.com/products/watsonx-ai) |  | | `mistral` | [Mistral](https://mistral.ai/) |  | | `moonshot_ai` | [Moonshot AI](https://www.moonshot.ai/) |  | | `openai` | [OpenAI](https://openai.com/) |  | | `perplexity` | [Perplexity](https://www.perplexity.ai/) |  | | `x_ai` | [xAI](https://x.ai/) |  | **[8]:** Used when accessing the 'generativelanguage.googleapis.com' endpoint. Also known as the AI Studio API. **[9]:** May be used when specific backend is unknown. **[10]:** Used when accessing the 'aiplatform.googleapis.com' endpoint. --- `gen_ai.token.modality` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `audio` | Audio tokens. |  | | `image` | Image tokens. |  | | `text` | Text tokens. |  | | `unknown` | The modality is not known. |  | ### Metric: `gen_ai.client.inference.usage.output_tokens` | Name | Instrument Type | Unit (UCUM) | Description | Stability | Entity Associations | | -------- | --------------- | ----------- | -------------- | --------- | ------ | | `gen_ai.client.inference.usage.output_tokens` | Counter | `{token}` | The number of output (completion) tokens used, including reasoning tokens. [1] |  | | **[1]:** This metric SHOULD be reported when an operation involves the usage of tokens and the count is readily available. When systems report both used tokens and billable tokens, instrumentation MUST report billable tokens. Instrumentation SHOULD record contributions from successful and failed inference operations. Each measurement MUST be a known, strictly positive modality contribution. When a contribution's modality cannot reliably be determined, instrumentation MUST set `gen_ai.token.modality` to `unknown`. **Requirement level:** [Recommended](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/general/signal-requirement-level.md). **Attributes:** | Key | Stability | [Requirement Level](https://opentelemetry.io/docs/specs/semconv/general/attribute-requirement-level/) | Value Type | Description | Example Values | | --- | --- | --- | --- | --- | --- | | [`gen_ai.operation.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The name of the operation being performed. [1] | `chat`; `generate_content`; `text_completion` | | [`gen_ai.provider.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The Generative AI provider as identified by the client or server instrumentation. [2] | `openai`; `gcp.gen_ai`; `gcp.vertex_ai` | | [`gen_ai.token.modality`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The modality of the tokens being counted. [3] | `text`; `image`; `audio` | | [`gen_ai.request.model`](/docs/registry/attributes/gen-ai.md) |  | `Conditionally Required` If available. | string | The name of the GenAI model a request is being made to. [4] | `gpt-4` | | [`server.port`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Conditionally Required` If `server.address` is set. | int | GenAI server port. [5] | `80`; `8080`; `443` | | [`gen_ai.response.model`](/docs/registry/attributes/gen-ai.md) |  | `Recommended` | string | The name of the model that generated the response. | `gpt-4-0613` | | [`server.address`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Recommended` | string | GenAI server address. [6] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | **[1] `gen_ai.operation.name`:** If one of the predefined values applies, but specific system uses a different name it's RECOMMENDED to document it in the semantic conventions for specific GenAI system and use system-specific name in the instrumentation. If a different name is not documented, instrumentation libraries SHOULD use applicable predefined value. **[2] `gen_ai.provider.name`:** Semantic conventions for individual GenAI operations SHOULD clarify which kinds of providers (e.g. inference, embeddings, retrieval, memory, hosted agent providers) apply when it is not clear from context. The attribute SHOULD be set based on the instrumentation's best knowledge and may differ from the actual upstream provider. For example, a client SDK may be configured against a proxy or hosting platform that transparently relays requests to a different provider. The `gen_ai.provider.name` attribute acts as a discriminator that identifies the GenAI telemetry format flavor specific to that provider within GenAI semantic conventions. It SHOULD be set consistently with provider-specific attributes and signals. For example, GenAI spans, metrics, and events related to AWS Bedrock should have the `gen_ai.provider.name` set to `aws.bedrock` and include applicable `aws.bedrock.*` attributes and are not expected to include `openai.*` attributes. **[3] `gen_ai.token.modality`:** When the provider does not break usage down by modality and the modality cannot be reliably determined, instrumentations SHOULD report the tokens under the `unknown` modality. **[4] `gen_ai.request.model`:** The name of the GenAI model a request is being made to. If the model is supplied by a vendor, then the value must be the exact name of the model requested. If the model is a fine-tuned custom model, the value should have a more specific name than the base model that's been fine-tuned. **[5] `server.port`:** When observed from the client side, and when communicating through an intermediary, `server.port` SHOULD represent the server port behind any intermediaries, for example proxies, if it's available. **[6] `server.address`:** When observed from the client side, and when communicating through an intermediary, `server.address` SHOULD represent the server address behind any intermediaries, for example proxies, if it's available. --- `gen_ai.operation.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `chat` | Chat completion operation such as [OpenAI Chat API](https://platform.openai.com/docs/api-reference/chat) |  | | `create_agent` | Create GenAI agent |  | | `create_memory` | Create new memory records |  | | `create_memory_store` | Create or initialize a memory store |  | | `delete_memory` | Delete memory records |  | | `delete_memory_store` | Delete or deprovision a memory store |  | | `embeddings` | Embeddings operation such as [OpenAI Create embeddings API](https://platform.openai.com/docs/api-reference/embeddings/create) |  | | `execute_tool` | Execute a tool |  | | `fetch_response` | Fetch a previously generated model response by its identifier, without performing inference, such as [OpenAI Get a model response](https://platform.openai.com/docs/api-reference/responses/get) [7] |  | | `generate_content` | Multimodal content generation operation such as [Gemini Generate Content](https://ai.google.dev/api/generate-content) |  | | `invoke_agent` | Invoke GenAI agent |  | | `invoke_workflow` | Invoke GenAI workflow |  | | `plan` | Agent planning or task decomposition phase |  | | `retrieval` | Retrieval operation such as [OpenAI Search Vector Store API](https://platform.openai.com/docs/api-reference/vector-stores/search) |  | | `search_memory` | Search/query memories from a memory store |  | | `text_completion` | Text completions operation such as [OpenAI Completions API (Legacy)](https://platform.openai.com/docs/api-reference/completions) |  | | `update_memory` | Update existing memory records |  | | `upsert_memory` | Create or update memory records without the caller choosing which |  | **[7]:** Instrumentations SHOULD NOT report token usage (as attributes or metrics) for this operation. --- `gen_ai.provider.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `anthropic` | [Anthropic](https://www.anthropic.com/) |  | | `aws.bedrock` | [AWS Bedrock](https://aws.amazon.com/bedrock) |  | | `azure.ai.inference` | Azure AI Inference |  | | `azure.ai.openai` | [Azure OpenAI](https://learn.microsoft.com/en-us/azure/ai-services/openai/overview) |  | | `cohere` | [Cohere](https://cohere.com/) |  | | `deepseek` | [DeepSeek](https://www.deepseek.com/) |  | | `gcp.gemini` | [Gemini](https://cloud.google.com/products/gemini) [8] |  | | `gcp.gen_ai` | Any Google generative AI endpoint [9] |  | | `gcp.vertex_ai` | [Vertex AI](https://cloud.google.com/vertex-ai) [10] |  | | `groq` | [Groq](https://groq.com/) |  | | `ibm.watsonx.ai` | [IBM Watsonx AI](https://www.ibm.com/products/watsonx-ai) |  | | `mistral` | [Mistral](https://mistral.ai/) |  | | `moonshot_ai` | [Moonshot AI](https://www.moonshot.ai/) |  | | `openai` | [OpenAI](https://openai.com/) |  | | `perplexity` | [Perplexity](https://www.perplexity.ai/) |  | | `x_ai` | [xAI](https://x.ai/) |  | **[8]:** Used when accessing the 'generativelanguage.googleapis.com' endpoint. Also known as the AI Studio API. **[9]:** May be used when specific backend is unknown. **[10]:** Used when accessing the 'aiplatform.googleapis.com' endpoint. --- `gen_ai.token.modality` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `audio` | Audio tokens. |  | | `image` | Image tokens. |  | | `text` | Text tokens. |  | | `unknown` | The modality is not known. |  | ### Metric: `gen_ai.client.inference.usage.cache_read.input_tokens` | Name | Instrument Type | Unit (UCUM) | Description | Stability | Entity Associations | | -------- | --------------- | ----------- | -------------- | --------- | ------ | | `gen_ai.client.inference.usage.cache_read.input_tokens` | Counter | `{token}` | The number of input tokens served from a provider-managed cache. [1] |  | | **[1]:** This value is a subset of `gen_ai.client.inference.usage.input_tokens`. Instrumentation SHOULD record contributions from successful and failed inference operations. Each measurement MUST be a known, strictly positive modality contribution. When a contribution's modality cannot reliably be determined, instrumentation MUST set `gen_ai.token.modality` to `unknown`. **Requirement level:** [Recommended](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/general/signal-requirement-level.md). **Attributes:** | Key | Stability | [Requirement Level](https://opentelemetry.io/docs/specs/semconv/general/attribute-requirement-level/) | Value Type | Description | Example Values | | --- | --- | --- | --- | --- | --- | | [`gen_ai.operation.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The name of the operation being performed. [1] | `chat`; `generate_content`; `text_completion` | | [`gen_ai.provider.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The Generative AI provider as identified by the client or server instrumentation. [2] | `openai`; `gcp.gen_ai`; `gcp.vertex_ai` | | [`gen_ai.token.modality`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The modality of the tokens being counted. [3] | `text`; `image`; `audio` | | [`gen_ai.request.model`](/docs/registry/attributes/gen-ai.md) |  | `Conditionally Required` If available. | string | The name of the GenAI model a request is being made to. [4] | `gpt-4` | | [`server.port`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Conditionally Required` If `server.address` is set. | int | GenAI server port. [5] | `80`; `8080`; `443` | | [`gen_ai.response.model`](/docs/registry/attributes/gen-ai.md) |  | `Recommended` | string | The name of the model that generated the response. | `gpt-4-0613` | | [`server.address`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Recommended` | string | GenAI server address. [6] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | **[1] `gen_ai.operation.name`:** If one of the predefined values applies, but specific system uses a different name it's RECOMMENDED to document it in the semantic conventions for specific GenAI system and use system-specific name in the instrumentation. If a different name is not documented, instrumentation libraries SHOULD use applicable predefined value. **[2] `gen_ai.provider.name`:** Semantic conventions for individual GenAI operations SHOULD clarify which kinds of providers (e.g. inference, embeddings, retrieval, memory, hosted agent providers) apply when it is not clear from context. The attribute SHOULD be set based on the instrumentation's best knowledge and may differ from the actual upstream provider. For example, a client SDK may be configured against a proxy or hosting platform that transparently relays requests to a different provider. The `gen_ai.provider.name` attribute acts as a discriminator that identifies the GenAI telemetry format flavor specific to that provider within GenAI semantic conventions. It SHOULD be set consistently with provider-specific attributes and signals. For example, GenAI spans, metrics, and events related to AWS Bedrock should have the `gen_ai.provider.name` set to `aws.bedrock` and include applicable `aws.bedrock.*` attributes and are not expected to include `openai.*` attributes. **[3] `gen_ai.token.modality`:** When the provider does not break usage down by modality and the modality cannot be reliably determined, instrumentations SHOULD report the tokens under the `unknown` modality. **[4] `gen_ai.request.model`:** The name of the GenAI model a request is being made to. If the model is supplied by a vendor, then the value must be the exact name of the model requested. If the model is a fine-tuned custom model, the value should have a more specific name than the base model that's been fine-tuned. **[5] `server.port`:** When observed from the client side, and when communicating through an intermediary, `server.port` SHOULD represent the server port behind any intermediaries, for example proxies, if it's available. **[6] `server.address`:** When observed from the client side, and when communicating through an intermediary, `server.address` SHOULD represent the server address behind any intermediaries, for example proxies, if it's available. --- `gen_ai.operation.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `chat` | Chat completion operation such as [OpenAI Chat API](https://platform.openai.com/docs/api-reference/chat) |  | | `create_agent` | Create GenAI agent |  | | `create_memory` | Create new memory records |  | | `create_memory_store` | Create or initialize a memory store |  | | `delete_memory` | Delete memory records |  | | `delete_memory_store` | Delete or deprovision a memory store |  | | `embeddings` | Embeddings operation such as [OpenAI Create embeddings API](https://platform.openai.com/docs/api-reference/embeddings/create) |  | | `execute_tool` | Execute a tool |  | | `fetch_response` | Fetch a previously generated model response by its identifier, without performing inference, such as [OpenAI Get a model response](https://platform.openai.com/docs/api-reference/responses/get) [7] |  | | `generate_content` | Multimodal content generation operation such as [Gemini Generate Content](https://ai.google.dev/api/generate-content) |  | | `invoke_agent` | Invoke GenAI agent |  | | `invoke_workflow` | Invoke GenAI workflow |  | | `plan` | Agent planning or task decomposition phase |  | | `retrieval` | Retrieval operation such as [OpenAI Search Vector Store API](https://platform.openai.com/docs/api-reference/vector-stores/search) |  | | `search_memory` | Search/query memories from a memory store |  | | `text_completion` | Text completions operation such as [OpenAI Completions API (Legacy)](https://platform.openai.com/docs/api-reference/completions) |  | | `update_memory` | Update existing memory records |  | | `upsert_memory` | Create or update memory records without the caller choosing which |  | **[7]:** Instrumentations SHOULD NOT report token usage (as attributes or metrics) for this operation. --- `gen_ai.provider.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `anthropic` | [Anthropic](https://www.anthropic.com/) |  | | `aws.bedrock` | [AWS Bedrock](https://aws.amazon.com/bedrock) |  | | `azure.ai.inference` | Azure AI Inference |  | | `azure.ai.openai` | [Azure OpenAI](https://learn.microsoft.com/en-us/azure/ai-services/openai/overview) |  | | `cohere` | [Cohere](https://cohere.com/) |  | | `deepseek` | [DeepSeek](https://www.deepseek.com/) |  | | `gcp.gemini` | [Gemini](https://cloud.google.com/products/gemini) [8] |  | | `gcp.gen_ai` | Any Google generative AI endpoint [9] |  | | `gcp.vertex_ai` | [Vertex AI](https://cloud.google.com/vertex-ai) [10] |  | | `groq` | [Groq](https://groq.com/) |  | | `ibm.watsonx.ai` | [IBM Watsonx AI](https://www.ibm.com/products/watsonx-ai) |  | | `mistral` | [Mistral](https://mistral.ai/) |  | | `moonshot_ai` | [Moonshot AI](https://www.moonshot.ai/) |  | | `openai` | [OpenAI](https://openai.com/) |  | | `perplexity` | [Perplexity](https://www.perplexity.ai/) |  | | `x_ai` | [xAI](https://x.ai/) |  | **[8]:** Used when accessing the 'generativelanguage.googleapis.com' endpoint. Also known as the AI Studio API. **[9]:** May be used when specific backend is unknown. **[10]:** Used when accessing the 'aiplatform.googleapis.com' endpoint. --- `gen_ai.token.modality` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `audio` | Audio tokens. |  | | `image` | Image tokens. |  | | `text` | Text tokens. |  | | `unknown` | The modality is not known. |  | ### Metric: `gen_ai.client.inference.usage.cache_write.input_tokens` | Name | Instrument Type | Unit (UCUM) | Description | Stability | Entity Associations | | -------- | --------------- | ----------- | -------------- | --------- | ------ | | `gen_ai.client.inference.usage.cache_write.input_tokens` | Counter | `{token}` | The number of input tokens written to a provider-managed cache. [1] |  | | **[1]:** This value is a subset of `gen_ai.client.inference.usage.input_tokens`. Instrumentation SHOULD record contributions from successful and failed inference operations. Each measurement MUST be a known, strictly positive modality contribution. When a contribution's modality cannot reliably be determined, instrumentation MUST set `gen_ai.token.modality` to `unknown`. **Requirement level:** [Recommended](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/general/signal-requirement-level.md). **Attributes:** | Key | Stability | [Requirement Level](https://opentelemetry.io/docs/specs/semconv/general/attribute-requirement-level/) | Value Type | Description | Example Values | | --- | --- | --- | --- | --- | --- | | [`gen_ai.operation.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The name of the operation being performed. [1] | `chat`; `generate_content`; `text_completion` | | [`gen_ai.provider.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The Generative AI provider as identified by the client or server instrumentation. [2] | `openai`; `gcp.gen_ai`; `gcp.vertex_ai` | | [`gen_ai.token.modality`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The modality of the tokens being counted. [3] | `text`; `image`; `audio` | | [`gen_ai.request.model`](/docs/registry/attributes/gen-ai.md) |  | `Conditionally Required` If available. | string | The name of the GenAI model a request is being made to. [4] | `gpt-4` | | [`server.port`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Conditionally Required` If `server.address` is set. | int | GenAI server port. [5] | `80`; `8080`; `443` | | [`gen_ai.response.model`](/docs/registry/attributes/gen-ai.md) |  | `Recommended` | string | The name of the model that generated the response. | `gpt-4-0613` | | [`server.address`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Recommended` | string | GenAI server address. [6] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | **[1] `gen_ai.operation.name`:** If one of the predefined values applies, but specific system uses a different name it's RECOMMENDED to document it in the semantic conventions for specific GenAI system and use system-specific name in the instrumentation. If a different name is not documented, instrumentation libraries SHOULD use applicable predefined value. **[2] `gen_ai.provider.name`:** Semantic conventions for individual GenAI operations SHOULD clarify which kinds of providers (e.g. inference, embeddings, retrieval, memory, hosted agent providers) apply when it is not clear from context. The attribute SHOULD be set based on the instrumentation's best knowledge and may differ from the actual upstream provider. For example, a client SDK may be configured against a proxy or hosting platform that transparently relays requests to a different provider. The `gen_ai.provider.name` attribute acts as a discriminator that identifies the GenAI telemetry format flavor specific to that provider within GenAI semantic conventions. It SHOULD be set consistently with provider-specific attributes and signals. For example, GenAI spans, metrics, and events related to AWS Bedrock should have the `gen_ai.provider.name` set to `aws.bedrock` and include applicable `aws.bedrock.*` attributes and are not expected to include `openai.*` attributes. **[3] `gen_ai.token.modality`:** When the provider does not break usage down by modality and the modality cannot be reliably determined, instrumentations SHOULD report the tokens under the `unknown` modality. **[4] `gen_ai.request.model`:** The name of the GenAI model a request is being made to. If the model is supplied by a vendor, then the value must be the exact name of the model requested. If the model is a fine-tuned custom model, the value should have a more specific name than the base model that's been fine-tuned. **[5] `server.port`:** When observed from the client side, and when communicating through an intermediary, `server.port` SHOULD represent the server port behind any intermediaries, for example proxies, if it's available. **[6] `server.address`:** When observed from the client side, and when communicating through an intermediary, `server.address` SHOULD represent the server address behind any intermediaries, for example proxies, if it's available. --- `gen_ai.operation.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `chat` | Chat completion operation such as [OpenAI Chat API](https://platform.openai.com/docs/api-reference/chat) |  | | `create_agent` | Create GenAI agent |  | | `create_memory` | Create new memory records |  | | `create_memory_store` | Create or initialize a memory store |  | | `delete_memory` | Delete memory records |  | | `delete_memory_store` | Delete or deprovision a memory store |  | | `embeddings` | Embeddings operation such as [OpenAI Create embeddings API](https://platform.openai.com/docs/api-reference/embeddings/create) |  | | `execute_tool` | Execute a tool |  | | `fetch_response` | Fetch a previously generated model response by its identifier, without performing inference, such as [OpenAI Get a model response](https://platform.openai.com/docs/api-reference/responses/get) [7] |  | | `generate_content` | Multimodal content generation operation such as [Gemini Generate Content](https://ai.google.dev/api/generate-content) |  | | `invoke_agent` | Invoke GenAI agent |  | | `invoke_workflow` | Invoke GenAI workflow |  | | `plan` | Agent planning or task decomposition phase |  | | `retrieval` | Retrieval operation such as [OpenAI Search Vector Store API](https://platform.openai.com/docs/api-reference/vector-stores/search) |  | | `search_memory` | Search/query memories from a memory store |  | | `text_completion` | Text completions operation such as [OpenAI Completions API (Legacy)](https://platform.openai.com/docs/api-reference/completions) |  | | `update_memory` | Update existing memory records |  | | `upsert_memory` | Create or update memory records without the caller choosing which |  | **[7]:** Instrumentations SHOULD NOT report token usage (as attributes or metrics) for this operation. --- `gen_ai.provider.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `anthropic` | [Anthropic](https://www.anthropic.com/) |  | | `aws.bedrock` | [AWS Bedrock](https://aws.amazon.com/bedrock) |  | | `azure.ai.inference` | Azure AI Inference |  | | `azure.ai.openai` | [Azure OpenAI](https://learn.microsoft.com/en-us/azure/ai-services/openai/overview) |  | | `cohere` | [Cohere](https://cohere.com/) |  | | `deepseek` | [DeepSeek](https://www.deepseek.com/) |  | | `gcp.gemini` | [Gemini](https://cloud.google.com/products/gemini) [8] |  | | `gcp.gen_ai` | Any Google generative AI endpoint [9] |  | | `gcp.vertex_ai` | [Vertex AI](https://cloud.google.com/vertex-ai) [10] |  | | `groq` | [Groq](https://groq.com/) |  | | `ibm.watsonx.ai` | [IBM Watsonx AI](https://www.ibm.com/products/watsonx-ai) |  | | `mistral` | [Mistral](https://mistral.ai/) |  | | `moonshot_ai` | [Moonshot AI](https://www.moonshot.ai/) |  | | `openai` | [OpenAI](https://openai.com/) |  | | `perplexity` | [Perplexity](https://www.perplexity.ai/) |  | | `x_ai` | [xAI](https://x.ai/) |  | **[8]:** Used when accessing the 'generativelanguage.googleapis.com' endpoint. Also known as the AI Studio API. **[9]:** May be used when specific backend is unknown. **[10]:** Used when accessing the 'aiplatform.googleapis.com' endpoint. --- `gen_ai.token.modality` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `audio` | Audio tokens. |  | | `image` | Image tokens. |  | | `text` | Text tokens. |  | | `unknown` | The modality is not known. |  | ### Metric: `gen_ai.client.inference.usage.reasoning.output_tokens` | Name | Instrument Type | Unit (UCUM) | Description | Stability | Entity Associations | | -------- | --------------- | ----------- | -------------- | --------- | ------ | | `gen_ai.client.inference.usage.reasoning.output_tokens` | Counter | `{token}` | The number of output tokens used for reasoning (e.g. chain-of-thought, extended thinking). [1] |  | | **[1]:** This value is a subset of `gen_ai.client.inference.usage.output_tokens`. Instrumentation SHOULD record contributions from successful and failed inference operations. Each measurement MUST be a known, strictly positive modality contribution. When a contribution's modality cannot reliably be determined, instrumentation MUST set `gen_ai.token.modality` to `unknown`. **Requirement level:** [Recommended](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/general/signal-requirement-level.md). **Attributes:** | Key | Stability | [Requirement Level](https://opentelemetry.io/docs/specs/semconv/general/attribute-requirement-level/) | Value Type | Description | Example Values | | --- | --- | --- | --- | --- | --- | | [`gen_ai.operation.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The name of the operation being performed. [1] | `chat`; `generate_content`; `text_completion` | | [`gen_ai.provider.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The Generative AI provider as identified by the client or server instrumentation. [2] | `openai`; `gcp.gen_ai`; `gcp.vertex_ai` | | [`gen_ai.token.modality`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The modality of the tokens being counted. [3] | `text`; `image`; `audio` | | [`gen_ai.request.model`](/docs/registry/attributes/gen-ai.md) |  | `Conditionally Required` If available. | string | The name of the GenAI model a request is being made to. [4] | `gpt-4` | | [`server.port`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Conditionally Required` If `server.address` is set. | int | GenAI server port. [5] | `80`; `8080`; `443` | | [`gen_ai.response.model`](/docs/registry/attributes/gen-ai.md) |  | `Recommended` | string | The name of the model that generated the response. | `gpt-4-0613` | | [`server.address`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Recommended` | string | GenAI server address. [6] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | **[1] `gen_ai.operation.name`:** If one of the predefined values applies, but specific system uses a different name it's RECOMMENDED to document it in the semantic conventions for specific GenAI system and use system-specific name in the instrumentation. If a different name is not documented, instrumentation libraries SHOULD use applicable predefined value. **[2] `gen_ai.provider.name`:** Semantic conventions for individual GenAI operations SHOULD clarify which kinds of providers (e.g. inference, embeddings, retrieval, memory, hosted agent providers) apply when it is not clear from context. The attribute SHOULD be set based on the instrumentation's best knowledge and may differ from the actual upstream provider. For example, a client SDK may be configured against a proxy or hosting platform that transparently relays requests to a different provider. The `gen_ai.provider.name` attribute acts as a discriminator that identifies the GenAI telemetry format flavor specific to that provider within GenAI semantic conventions. It SHOULD be set consistently with provider-specific attributes and signals. For example, GenAI spans, metrics, and events related to AWS Bedrock should have the `gen_ai.provider.name` set to `aws.bedrock` and include applicable `aws.bedrock.*` attributes and are not expected to include `openai.*` attributes. **[3] `gen_ai.token.modality`:** When the provider does not break usage down by modality and the modality cannot be reliably determined, instrumentations SHOULD report the tokens under the `unknown` modality. **[4] `gen_ai.request.model`:** The name of the GenAI model a request is being made to. If the model is supplied by a vendor, then the value must be the exact name of the model requested. If the model is a fine-tuned custom model, the value should have a more specific name than the base model that's been fine-tuned. **[5] `server.port`:** When observed from the client side, and when communicating through an intermediary, `server.port` SHOULD represent the server port behind any intermediaries, for example proxies, if it's available. **[6] `server.address`:** When observed from the client side, and when communicating through an intermediary, `server.address` SHOULD represent the server address behind any intermediaries, for example proxies, if it's available. --- `gen_ai.operation.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `chat` | Chat completion operation such as [OpenAI Chat API](https://platform.openai.com/docs/api-reference/chat) |  | | `create_agent` | Create GenAI agent |  | | `create_memory` | Create new memory records |  | | `create_memory_store` | Create or initialize a memory store |  | | `delete_memory` | Delete memory records |  | | `delete_memory_store` | Delete or deprovision a memory store |  | | `embeddings` | Embeddings operation such as [OpenAI Create embeddings API](https://platform.openai.com/docs/api-reference/embeddings/create) |  | | `execute_tool` | Execute a tool |  | | `fetch_response` | Fetch a previously generated model response by its identifier, without performing inference, such as [OpenAI Get a model response](https://platform.openai.com/docs/api-reference/responses/get) [7] |  | | `generate_content` | Multimodal content generation operation such as [Gemini Generate Content](https://ai.google.dev/api/generate-content) |  | | `invoke_agent` | Invoke GenAI agent |  | | `invoke_workflow` | Invoke GenAI workflow |  | | `plan` | Agent planning or task decomposition phase |  | | `retrieval` | Retrieval operation such as [OpenAI Search Vector Store API](https://platform.openai.com/docs/api-reference/vector-stores/search) |  | | `search_memory` | Search/query memories from a memory store |  | | `text_completion` | Text completions operation such as [OpenAI Completions API (Legacy)](https://platform.openai.com/docs/api-reference/completions) |  | | `update_memory` | Update existing memory records |  | | `upsert_memory` | Create or update memory records without the caller choosing which |  | **[7]:** Instrumentations SHOULD NOT report token usage (as attributes or metrics) for this operation. --- `gen_ai.provider.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `anthropic` | [Anthropic](https://www.anthropic.com/) |  | | `aws.bedrock` | [AWS Bedrock](https://aws.amazon.com/bedrock) |  | | `azure.ai.inference` | Azure AI Inference |  | | `azure.ai.openai` | [Azure OpenAI](https://learn.microsoft.com/en-us/azure/ai-services/openai/overview) |  | | `cohere` | [Cohere](https://cohere.com/) |  | | `deepseek` | [DeepSeek](https://www.deepseek.com/) |  | | `gcp.gemini` | [Gemini](https://cloud.google.com/products/gemini) [8] |  | | `gcp.gen_ai` | Any Google generative AI endpoint [9] |  | | `gcp.vertex_ai` | [Vertex AI](https://cloud.google.com/vertex-ai) [10] |  | | `groq` | [Groq](https://groq.com/) |  | | `ibm.watsonx.ai` | [IBM Watsonx AI](https://www.ibm.com/products/watsonx-ai) |  | | `mistral` | [Mistral](https://mistral.ai/) |  | | `moonshot_ai` | [Moonshot AI](https://www.moonshot.ai/) |  | | `openai` | [OpenAI](https://openai.com/) |  | | `perplexity` | [Perplexity](https://www.perplexity.ai/) |  | | `x_ai` | [xAI](https://x.ai/) |  | **[8]:** Used when accessing the 'generativelanguage.googleapis.com' endpoint. Also known as the AI Studio API. **[9]:** May be used when specific backend is unknown. **[10]:** Used when accessing the 'aiplatform.googleapis.com' endpoint. --- `gen_ai.token.modality` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `audio` | Audio tokens. |  | | `image` | Image tokens. |  | | `text` | Text tokens. |  | | `unknown` | The modality is not known. |  | ### Metric: `gen_ai.client.inference.operation.input_tokens` This metric SHOULD be specified with [ExplicitBucketBoundaries] of [1, 4, 16, 64, 256, 1024, 4096, 16384, 65536, 262144, 1048576, 4194304, 16777216, 67108864]. | Name | Instrument Type | Unit (UCUM) | Description | Stability | Entity Associations | | -------- | --------------- | ----------- | -------------- | --------- | ------ | | `gen_ai.client.inference.operation.input_tokens` | Histogram | `{token}` | The number of input (prompt) tokens used per inference operation. [1] |  | | **[1]:** This metric SHOULD be reported when an operation involves the usage of tokens and the count is readily available. For example, if GenAI system returns usage information in the streaming response, it SHOULD be used. Or if GenAI system returns each token independently, instrumentation SHOULD count number of output tokens and record the result. If instrumentation cannot efficiently obtain number of input and/or output tokens, it MAY allow users to enable offline token counting. Otherwise it MUST NOT report usage metrics. When systems report both used tokens and billable tokens, instrumentation MUST report billable tokens. Instrumentations SHOULD record this metric for each inference operation, including failed ones. The recorded value SHOULD match the token count returned by the provider, or `0` if the response was not received or did not contain token usage. **Requirement level:** [Recommended](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/general/signal-requirement-level.md). **Attributes:** | Key | Stability | [Requirement Level](https://opentelemetry.io/docs/specs/semconv/general/attribute-requirement-level/) | Value Type | Description | Example Values | | --- | --- | --- | --- | --- | --- | | [`gen_ai.operation.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The name of the operation being performed. [1] | `chat`; `generate_content`; `text_completion` | | [`gen_ai.provider.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The Generative AI provider as identified by the client or server instrumentation. [2] | `openai`; `gcp.gen_ai`; `gcp.vertex_ai` | | [`error.type`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/error.md) |  | `Conditionally Required` If the operation ended in an error. | string | Describes a class of error the operation ended with. [3] | `timeout`; `java.net.UnknownHostException`; `server_certificate_invalid`; `500` | | [`gen_ai.request.model`](/docs/registry/attributes/gen-ai.md) |  | `Conditionally Required` If available. | string | The name of the GenAI model a request is being made to. [4] | `gpt-4` | | [`server.port`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Conditionally Required` If `server.address` is set. | int | GenAI server port. [5] | `80`; `8080`; `443` | | [`gen_ai.response.model`](/docs/registry/attributes/gen-ai.md) |  | `Recommended` | string | The name of the model that generated the response. | `gpt-4-0613` | | [`server.address`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Recommended` | string | GenAI server address. [6] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | **[1] `gen_ai.operation.name`:** If one of the predefined values applies, but specific system uses a different name it's RECOMMENDED to document it in the semantic conventions for specific GenAI system and use system-specific name in the instrumentation. If a different name is not documented, instrumentation libraries SHOULD use applicable predefined value. **[2] `gen_ai.provider.name`:** Semantic conventions for individual GenAI operations SHOULD clarify which kinds of providers (e.g. inference, embeddings, retrieval, memory, hosted agent providers) apply when it is not clear from context. The attribute SHOULD be set based on the instrumentation's best knowledge and may differ from the actual upstream provider. For example, a client SDK may be configured against a proxy or hosting platform that transparently relays requests to a different provider. The `gen_ai.provider.name` attribute acts as a discriminator that identifies the GenAI telemetry format flavor specific to that provider within GenAI semantic conventions. It SHOULD be set consistently with provider-specific attributes and signals. For example, GenAI spans, metrics, and events related to AWS Bedrock should have the `gen_ai.provider.name` set to `aws.bedrock` and include applicable `aws.bedrock.*` attributes and are not expected to include `openai.*` attributes. **[3] `error.type`:** The `error.type` SHOULD match the error code returned by the Generative AI provider or the client library, the canonical name of exception that occurred, or another low-cardinality error identifier. Instrumentations SHOULD document the list of errors they report. **[4] `gen_ai.request.model`:** The name of the GenAI model a request is being made to. If the model is supplied by a vendor, then the value must be the exact name of the model requested. If the model is a fine-tuned custom model, the value should have a more specific name than the base model that's been fine-tuned. **[5] `server.port`:** When observed from the client side, and when communicating through an intermediary, `server.port` SHOULD represent the server port behind any intermediaries, for example proxies, if it's available. **[6] `server.address`:** When observed from the client side, and when communicating through an intermediary, `server.address` SHOULD represent the server address behind any intermediaries, for example proxies, if it's available. --- `error.type` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `_OTHER` | A fallback error value to be used when the instrumentation doesn't define a custom value. |  | --- `gen_ai.operation.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `chat` | Chat completion operation such as [OpenAI Chat API](https://platform.openai.com/docs/api-reference/chat) |  | | `create_agent` | Create GenAI agent |  | | `create_memory` | Create new memory records |  | | `create_memory_store` | Create or initialize a memory store |  | | `delete_memory` | Delete memory records |  | | `delete_memory_store` | Delete or deprovision a memory store |  | | `embeddings` | Embeddings operation such as [OpenAI Create embeddings API](https://platform.openai.com/docs/api-reference/embeddings/create) |  | | `execute_tool` | Execute a tool |  | | `fetch_response` | Fetch a previously generated model response by its identifier, without performing inference, such as [OpenAI Get a model response](https://platform.openai.com/docs/api-reference/responses/get) [7] |  | | `generate_content` | Multimodal content generation operation such as [Gemini Generate Content](https://ai.google.dev/api/generate-content) |  | | `invoke_agent` | Invoke GenAI agent |  | | `invoke_workflow` | Invoke GenAI workflow |  | | `plan` | Agent planning or task decomposition phase |  | | `retrieval` | Retrieval operation such as [OpenAI Search Vector Store API](https://platform.openai.com/docs/api-reference/vector-stores/search) |  | | `search_memory` | Search/query memories from a memory store |  | | `text_completion` | Text completions operation such as [OpenAI Completions API (Legacy)](https://platform.openai.com/docs/api-reference/completions) |  | | `update_memory` | Update existing memory records |  | | `upsert_memory` | Create or update memory records without the caller choosing which |  | **[7]:** Instrumentations SHOULD NOT report token usage (as attributes or metrics) for this operation. --- `gen_ai.provider.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `anthropic` | [Anthropic](https://www.anthropic.com/) |  | | `aws.bedrock` | [AWS Bedrock](https://aws.amazon.com/bedrock) |  | | `azure.ai.inference` | Azure AI Inference |  | | `azure.ai.openai` | [Azure OpenAI](https://learn.microsoft.com/en-us/azure/ai-services/openai/overview) |  | | `cohere` | [Cohere](https://cohere.com/) |  | | `deepseek` | [DeepSeek](https://www.deepseek.com/) |  | | `gcp.gemini` | [Gemini](https://cloud.google.com/products/gemini) [8] |  | | `gcp.gen_ai` | Any Google generative AI endpoint [9] |  | | `gcp.vertex_ai` | [Vertex AI](https://cloud.google.com/vertex-ai) [10] |  | | `groq` | [Groq](https://groq.com/) |  | | `ibm.watsonx.ai` | [IBM Watsonx AI](https://www.ibm.com/products/watsonx-ai) |  | | `mistral` | [Mistral](https://mistral.ai/) |  | | `moonshot_ai` | [Moonshot AI](https://www.moonshot.ai/) |  | | `openai` | [OpenAI](https://openai.com/) |  | | `perplexity` | [Perplexity](https://www.perplexity.ai/) |  | | `x_ai` | [xAI](https://x.ai/) |  | **[8]:** Used when accessing the 'generativelanguage.googleapis.com' endpoint. Also known as the AI Studio API. **[9]:** May be used when specific backend is unknown. **[10]:** Used when accessing the 'aiplatform.googleapis.com' endpoint. ### Metric: `gen_ai.client.inference.operation.output_tokens` This metric SHOULD be specified with [ExplicitBucketBoundaries] of [1, 4, 16, 64, 256, 1024, 4096, 16384, 65536, 262144, 1048576, 4194304, 16777216, 67108864]. | Name | Instrument Type | Unit (UCUM) | Description | Stability | Entity Associations | | -------- | --------------- | ----------- | -------------- | --------- | ------ | | `gen_ai.client.inference.operation.output_tokens` | Histogram | `{token}` | The number of output (completion) tokens used per inference operation. [1] |  | | **[1]:** This metric SHOULD be reported when an operation involves the usage of tokens and the count is readily available. For example, if GenAI system returns usage information in the streaming response, it SHOULD be used. Or if GenAI system returns each token independently, instrumentation SHOULD count number of output tokens and record the result. If instrumentation cannot efficiently obtain number of input and/or output tokens, it MAY allow users to enable offline token counting. Otherwise it MUST NOT report usage metrics. When systems report both used tokens and billable tokens, instrumentation MUST report billable tokens. Instrumentations SHOULD record this metric for each inference operation, including failed ones. The recorded value SHOULD match the token count returned by the provider, or `0` if the response was not received or did not contain token usage. **Requirement level:** [Recommended](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/general/signal-requirement-level.md). **Attributes:** | Key | Stability | [Requirement Level](https://opentelemetry.io/docs/specs/semconv/general/attribute-requirement-level/) | Value Type | Description | Example Values | | --- | --- | --- | --- | --- | --- | | [`gen_ai.operation.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The name of the operation being performed. [1] | `chat`; `generate_content`; `text_completion` | | [`gen_ai.provider.name`](/docs/registry/attributes/gen-ai.md) |  | `Required` | string | The Generative AI provider as identified by the client or server instrumentation. [2] | `openai`; `gcp.gen_ai`; `gcp.vertex_ai` | | [`error.type`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/error.md) |  | `Conditionally Required` If the operation ended in an error. | string | Describes a class of error the operation ended with. [3] | `timeout`; `java.net.UnknownHostException`; `server_certificate_invalid`; `500` | | [`gen_ai.request.model`](/docs/registry/attributes/gen-ai.md) |  | `Conditionally Required` If available. | string | The name of the GenAI model a request is being made to. [4] | `gpt-4` | | [`server.port`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Conditionally Required` If `server.address` is set. | int | GenAI server port. [5] | `80`; `8080`; `443` | | [`gen_ai.response.model`](/docs/registry/attributes/gen-ai.md) |  | `Recommended` | string | The name of the model that generated the response. | `gpt-4-0613` | | [`server.address`](https://github.com/open-telemetry/semantic-conventions/blob/v1.44.0/docs/registry/attributes/server.md) |  | `Recommended` | string | GenAI server address. [6] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | **[1] `gen_ai.operation.name`:** If one of the predefined values applies, but specific system uses a different name it's RECOMMENDED to document it in the semantic conventions for specific GenAI system and use system-specific name in the instrumentation. If a different name is not documented, instrumentation libraries SHOULD use applicable predefined value. **[2] `gen_ai.provider.name`:** Semantic conventions for individual GenAI operations SHOULD clarify which kinds of providers (e.g. inference, embeddings, retrieval, memory, hosted agent providers) apply when it is not clear from context. The attribute SHOULD be set based on the instrumentation's best knowledge and may differ from the actual upstream provider. For example, a client SDK may be configured against a proxy or hosting platform that transparently relays requests to a different provider. The `gen_ai.provider.name` attribute acts as a discriminator that identifies the GenAI telemetry format flavor specific to that provider within GenAI semantic conventions. It SHOULD be set consistently with provider-specific attributes and signals. For example, GenAI spans, metrics, and events related to AWS Bedrock should have the `gen_ai.provider.name` set to `aws.bedrock` and include applicable `aws.bedrock.*` attributes and are not expected to include `openai.*` attributes. **[3] `error.type`:** The `error.type` SHOULD match the error code returned by the Generative AI provider or the client library, the canonical name of exception that occurred, or another low-cardinality error identifier. Instrumentations SHOULD document the list of errors they report. **[4] `gen_ai.request.model`:** The name of the GenAI model a request is being made to. If the model is supplied by a vendor, then the value must be the exact name of the model requested. If the model is a fine-tuned custom model, the value should have a more specific name than the base model that's been fine-tuned. **[5] `server.port`:** When observed from the client side, and when communicating through an intermediary, `server.port` SHOULD represent the server port behind any intermediaries, for example proxies, if it's available. **[6] `server.address`:** When observed from the client side, and when communicating through an intermediary, `server.address` SHOULD represent the server address behind any intermediaries, for example proxies, if it's available. --- `error.type` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `_OTHER` | A fallback error value to be used when the instrumentation doesn't define a custom value. |  | --- `gen_ai.operation.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `chat` | Chat completion operation such as [OpenAI Chat API](https://platform.openai.com/docs/api-reference/chat) |  | | `create_agent` | Create GenAI agent |  | | `create_memory` | Create new memory records |  | | `create_memory_store` | Create or initialize a memory store |  | | `delete_memory` | Delete memory records |  | | `delete_memory_store` | Delete or deprovision a memory store |  | | `embeddings` | Embeddings operation such as [OpenAI Create embeddings API](https://platform.openai.com/docs/api-reference/embeddings/create) |  | | `execute_tool` | Execute a tool |  | | `fetch_response` | Fetch a previously generated model response by its identifier, without performing inference, such as [OpenAI Get a model response](https://platform.openai.com/docs/api-reference/responses/get) [7] |  | | `generate_content` | Multimodal content generation operation such as [Gemini Generate Content](https://ai.google.dev/api/generate-content) |  | | `invoke_agent` | Invoke GenAI agent |  | | `invoke_workflow` | Invoke GenAI workflow |  | | `plan` | Agent planning or task decomposition phase |  | | `retrieval` | Retrieval operation such as [OpenAI Search Vector Store API](https://platform.openai.com/docs/api-reference/vector-stores/search) |  | | `search_memory` | Search/query memories from a memory store |  | | `text_completion` | Text completions operation such as [OpenAI Completions API (Legacy)](https://platform.openai.com/docs/api-reference/completions) |  | | `update_memory` | Update existing memory records |  | | `upsert_memory` | Create or update memory records without the caller choosing which |  | **[7]:** Instrumentations SHOULD NOT report token usage (as attributes or metrics) for this operation. --- `gen_ai.provider.name` has the following list of well-known values. If one of them applies, then the respective value MUST be used; otherwise, a custom value MAY be used. | Value | Description | Stability | | --- | --- | --- | | `anthropic` | [Anthropic](https://www.anthropic.com/) |  | | `aws.bedrock` | [AWS Bedrock](https://aws.amazon.com/bedrock) |  | | `azure.ai.inference` | Azure AI Inference |  | | `azure.ai.openai` | [Azure OpenAI](https://learn.microsoft.com/en-us/azure/ai-services/openai/overview) |  | | `cohere` | [Cohere](https://cohere.com/) |  | | `deepseek` | [DeepSeek](https://www.deepseek.com/) |  | | `gcp.gemini` | [Gemini](https://cloud.google.com/products/gemini) [8] |  | | `gcp.gen_ai` | Any Google generative AI endpoint [9] |  | | `gcp.vertex_ai` | [Vertex AI](https://cloud.google.com/vertex-ai) [10] |  | | `groq` | [Groq](https://groq.com/) |  | | `ibm.watsonx.ai` | [IBM Watsonx AI](https://www.ibm.com/products/watsonx-ai) |  | | `mistral` | [Mistral](https://mistral.ai/) |  | | `moonshot_ai` | [Moonshot AI](https://www.moonshot.ai/) |  | | `openai` | [OpenAI](https://openai.com/) |  | | `perplexity` | [Perplexity](https://www.perplexity.ai/) |  | | `x_ai` | [xAI](https://x.ai/) |  | **[8]:** Used when accessing the 'generativelanguage.googleapis.com' endpoint. Also known as the AI Studio API. **[9]:** May be used when specific backend is unknown. **[10]:** Used when accessing the 'aiplatform.googleapis.com' endpoint. [DocumentStatus]: https://opentelemetry.io/docs/specs/otel/document-status [ExplicitBucketBoundaries]: https://github.com/open-telemetry/opentelemetry-specification/blob/v1.55.0/specification/metrics/api.md#instrument-advisory-parameters