generated: '2026-08-02' method: searched source: >- https://support.positron.ai/release-notes, https://support.positron.ai/api-documentation, https://www.positron.ai/atlas, openapi/positron-inference-openapi.yml versioning: scheme: uri-path current: v1 api_version: '1.0' docs: https://support.positron.ai/api-documentation evidence: >- inference OpenAPI server template `{host}/api/v1`; info.version '1.0' on both the Olivaw OpenAI API and the Olivaw Admin API documents. product_versioning: scheme: appliance release train current: Atlas v1.1 docs: https://support.positron.ai/release-notes note: >- The published release notes version the Atlas appliance / Olivaw software image, not the HTTP API. Atlas v1.1 added two quantized Llama 3.1 models (NeuralMagic GPTQ weights). deprecation: policy_url: null policy_published: false sunset_header: false deprecation_header: false note: >- No deprecation or sunset policy is published, and no RFC 8594 Sunset/Deprecation header support is documented. No operation in either OpenAPI is marked `deprecated: true`. sla: url: https://www.positron.ai/atlas uptime_target: null support_response: 24-hour SLA support from a US-based team (published on the Atlas product page) note: >- The SLA published is a hardware/support SLA for the Atlas appliance, not an API uptime commitment. No API availability target is published. status_page: null status_page_note: >- No status page. status.positron.ai does not resolve to a distinct host (wildcard DNS returns the apex) and does not answer HTTPS; no third-party status page (Statuspage/Instatus/BetterStack) was found for Positron AI. Consistent with an appliance/private-cluster deployment model where availability is the customer's own hardware. changelog: https://support.positron.ai/release-notes deprecated_operations: [] support: contact: https://www.positron.ai/contact-sales developer_portal: https://support.positron.ai/ api_contact_email: api@positron.ai