# Pricing transcribed from the provider's own page — see provenance. Numbers are quoted, not derived. generated: '2026-10-07' method: searched source: https://oumi.ai/pricing sources: - https://oumi.ai/pricing provenance: tool: all/0-working/economics/harvest_pricing.py run: 20261007T104522 grounding: every price_text was found verbatim in the fetched page text audience: developer-api pricing_basis: per-call delivery: delivery: managed+licensed method: page read: '2026-10-07' evidence: BYOC - VPC & on-prem managed_tier: Pro published: true note: Tiered plans plus detailed per-token, per-judgment, per-row, per-GB and per-GPU-hour usage rates; large inference model price table also listed (per 1M input/output tokens). FAQ mentions 'Is Oumi open source?' but no license stated. plan_count: 20 currency: USD plans: - name: Free price_text: $0 price: 0 currency: USD period: month included_quota: Up to $50 in credits for 3 months - name: Pro price_text: $25/month price: 25 currency: USD period: month included_quota: '$25/month credits platform-wide; Download model weights: 1 / month' overage: pay-as-you-go after - name: Enterprise price_text: Custom period: none included_quota: Everything in Pro; BYOC - VPC & on-prem - name: Oumi Agent tokens - Input price_text: $6.25 price: 6.25 currency: USD period: per-use unit: per 1M input tokens - name: Oumi Agent tokens - Output price_text: $31.25 price: 31.25 currency: USD period: per-use unit: per 1M output tokens - name: Oumi Agent tokens - Cache write price_text: $7.80 price: 7.8 currency: USD period: per-use unit: per 1M tokens - name: Oumi Agent tokens - Cache read price_text: $0.65 price: 0.65 currency: USD period: per-use unit: per 1M tokens - name: Evaluation price_text: 1,000 judgments / $1 currency: USD period: per-use unit: judgments - name: Data Synthesis price_text: 1,000 rows / $1 currency: USD period: per-use unit: rows - name: Storage price_text: 4 GB/month / $1 currency: USD period: month unit: GB - name: SFT up to 16B price_text: $0.49 price: 0.49 currency: USD period: per-use unit: per 1M training tokens - name: SFT 16.1–32B price_text: $2.00 price: 2 currency: USD period: per-use unit: per 1M training tokens - name: SFT 32.1–80B price_text: $3.00 price: 3 currency: USD period: per-use unit: per 1M training tokens - name: SFT 80.1–300B price_text: $6.00 price: 6 currency: USD period: per-use unit: per 1M training tokens - name: On-Policy Distillation A100-80GB price_text: $2.90 price: 2.9 currency: USD period: per-use unit: GPU-hr - name: On-Policy Distillation H100-80GB price_text: $4.00 price: 4 currency: USD period: per-use unit: GPU-hr - name: Production Inference A100-80GB price_text: $3.00 price: 3 currency: USD period: per-use unit: GPU-hr - name: Production Inference H100-80GB price_text: $7.00 price: 7 currency: USD period: per-use unit: GPU-hr - name: Production Inference H200-141GB price_text: $7.00 price: 7 currency: USD period: per-use unit: GPU-hr - name: Production Inference B200-180GB price_text: $10.00 price: 10 currency: USD period: per-use unit: GPU-hr x-canonical: generated: '2026-10-08' vocabulary: all/0-working/economics/plans-canonical.yml models: - free - hybrid - metered - subscription units: - data - token plans: - name: Free access: other model: free model_how: derived units: [] periods: - month - name: Pro access: other model: hybrid model_how: derived units: [] periods: - month - name: Enterprise access: other model: unstated model_how: derived units: [] periods: - none - name: Oumi Agent tokens - Input access: other model: hybrid model_how: derived units: - token periods: - other - name: Oumi Agent tokens - Output access: other model: hybrid model_how: derived units: - token periods: - other - name: Oumi Agent tokens - Cache write access: other model: hybrid model_how: derived units: - token periods: - other - name: Oumi Agent tokens - Cache read access: other model: hybrid model_how: derived units: - token periods: - other - name: Evaluation access: other model: unstated model_how: derived units: - other periods: - other - name: Data Synthesis access: other model: metered model_how: derived units: - data periods: - other - name: Storage access: other model: metered model_how: derived units: - data periods: - month - name: SFT up to 16B access: other model: hybrid model_how: derived units: - token periods: - other - name: SFT 16.1–32B access: other model: hybrid model_how: derived units: - token periods: - other - name: SFT 32.1–80B access: other model: hybrid model_how: derived units: - token periods: - other - name: SFT 80.1–300B access: other model: hybrid model_how: derived units: - token periods: - other - name: On-Policy Distillation A100-80GB access: other model: subscription model_how: derived units: - other periods: - other - name: On-Policy Distillation H100-80GB access: other model: subscription model_how: derived units: - other periods: - other - name: Production Inference A100-80GB access: other model: subscription model_how: derived units: - other periods: - other - name: Production Inference H100-80GB access: other model: subscription model_how: derived units: - other periods: - other - name: Production Inference H200-141GB access: other model: subscription model_how: derived units: - other periods: - other - name: Production Inference B200-180GB access: other model: subscription model_how: derived units: - other periods: - other