Token prices
What every current model from the six labs costs per million tokens, in and out, with the context window and output cap next to it. Refreshed three times a day; every change is logged.
As of Oct 19, 2026 · source OpenRouter model list · 133 models
Cheapest output$0.08/MLlama 3.1 8B Instruct · Meta
Priciest output$600/Mo1-pro · OpenAI
Largest context2MGrok 4.20 Multi-Agent
Last 120 days67 cuts · 78 hikes14 new listings
Cost calculator set your monthly volume, the table fills in what each model would cost
Presets:
133 models
| Model | Input$/M | Output$/M | Cached$/M | Batchin / out | Context | Max out | Listed | Monthlyyour volume |
|---|---|---|---|---|---|---|---|---|
| Claude Haiku 5.5Anthropic · reasoning · vision | $0.1 | $0.5 | $0.01 | $0.05 / $0.25 | 1M | 128K | ||
| Claude Sonnet 5.5Anthropic · reasoning · vision · 1 change | $2.00 | $10.0 | $0.1 | $1.00 / $5.00 | 1M | 128K | ||
| Claude Opus 5.5Anthropic · reasoning · vision | $4.00 | $20.0 | $0.2 | $2.00 / $10.0 | 1M | 128K | ||
| Claude Fable 5.1Anthropic · reasoning · vision | $10.0 | $50.0 | $0.25 | $5.00 / $25.0 | 1M | 128K | ||
| Claude Opus 5Anthropic · reasoning · vision | $5.00 | $25.0 | $0.5 | $2.50 / $12.5 | 1M | 128K | ||
| Claude Sonnet 5Anthropic · reasoning · vision | $2.00 | $10.0 | $0.2 | $1.00 / $5.00 | 1M | 128K | ||
| Claude Fable 5Anthropic · reasoning · vision | $10.0 | $50.0 | $1.00 | $5.00 / $25.0 | 1M | 128K | ||
| Claude Opus 4.8Anthropic · reasoning · vision | $5.00 | $25.0 | $0.5 | $2.50 / $12.5 | 1M | 128K | ||
| Claude Opus 4.7Anthropic · reasoning · vision | $5.00 | $25.0 | $0.5 | $2.50 / $12.5 | 1M | 128K | ||
| Claude Sonnet 4.6Anthropic · reasoning · vision | $3.00 | $15.0 | $0.3 | $1.50 / $7.50 | 1M | 128K | ||
| Claude Opus 4.6Anthropic · reasoning · vision | $5.00 | $25.0 | $0.5 | $2.50 / $12.5 | 1M | 128K | ||
| Claude Opus 4.5Anthropic · reasoning · vision | $5.00 | $25.0 | $0.5 | $2.50 / $12.5 | 200K | 64K | ||
| Claude Haiku 4.5Anthropic · reasoning · vision | $1.00 | $5.00 | $0.1 | $0.5 / $2.50 | 200K | 64K | ||
| Claude Sonnet 4.5Anthropic · reasoning · vision | $3.00 | $15.0 | $0.3 | $1.50 / $7.50 | 1M | 64K | ||
| Claude Opus 4.1Anthropic · reasoning · vision | $15.0 | $75.0 | $1.50 | $7.50 / $37.5 | 200K | 32K | ||
| Gemini 4 ArgonGoogle · reasoning · vision · 1 change | $2.00 | $10.0 | $0.1 | $1.00 / $5.00 | 1.0M | 944K | ||
| Nano Banana 2.1Google · reasoning · vision | $1.50 | $7.50 | · | · | 66K | 59K | ||
| Gemini 3.8 FlashGoogle · reasoning · vision · 1 change | $0.825 | $4.13 | $0.083 | $0.375 / $1.88 | 1.0M | 66K | ||
| Gemini 3.7 FlashGoogle · reasoning · vision | $0.75 | $3.75 | $0.075 | $0.375 / $1.88 | 1.0M | 66K | ||
| Gemini 3.6 FlashGoogle · reasoning · vision | $0.75 | $3.75 | $0.075 | $0.375 / $1.88 | 1.0M | 66K | ||
| Gemini 3.5 Flash LiteGoogle · reasoning · vision | $0.3 | $2.50 | $0.03 | $0.15 / $1.25 | 1.0M | 66K | ||
| Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)Google · reasoning · vision | $0.25 | $1.50 | · | · | 66K | 59K | ||
| Nano Banana 2 (Gemini 3.1 Flash Image)Google · reasoning · vision | $0.5 | $3.00 | · | · | 131K | 33K | ||
| Nano Banana Pro (Gemini 3 Pro Image)Google · reasoning · vision | $2.00 | $12.0 | $0.2 | · | 131K | 33K | ||
| Gemini 3.5 FlashGoogle · reasoning · vision | $1.50 | $9.00 | $0.15 | $0.75 / $4.50 | 1.0M | 66K | ||
| Gemini 3.1 Flash LiteGoogle · reasoning · vision | $0.25 | $1.50 | $0.025 | $0.125 / $0.75 | 1.0M | 66K | ||
| Gemma 4 26B A4B Google · reasoning · vision · 34 changes | $0.09 | $0.3 | $0.05 | · | 262K | 236K | ||
| Gemma 4 31BGoogle · reasoning · vision · 1 change | $0.2 | $0.4 | · | · | 262K | 16K | ||
| Lyria 3 Pro PreviewGoogle · vision | $0.0 | $0.0 | · | · | 1.0M | 66K | ||
| Lyria 3 Clip PreviewGoogle · vision | $0.0 | $0.0 | · | · | 1.0M | 66K | ||
| Gemini 3.1 Flash Lite PreviewGoogle · reasoning · vision | $0.25 | $1.50 | $0.025 | · | 1.0M | 66K | ||
| Nano Banana 2 (Gemini 3.1 Flash Image Preview)Google · reasoning · vision | $0.5 | $3.00 | · | · | 66K | 59K | ||
| Gemini 3.1 Pro Preview Custom ToolsGoogle · reasoning · vision | $2.00 | $12.0 | $0.2 | · | 1.0M | 66K | ||
| Gemini 3.1 Pro PreviewGoogle · reasoning · vision | $2.00 | $12.0 | $0.2 | $1.00 / $6.00 | 1.0M | 66K | ||
| Gemini 3 Flash PreviewGoogle · reasoning · vision | $0.5 | $3.00 | $0.05 | $0.25 / $1.50 | 1.0M | 66K | ||
| Nano Banana Pro (Gemini 3 Pro Image Preview)Google · reasoning · vision | $2.00 | $12.0 | $0.2 | · | 66K | 33K | ||
| Nano Banana (Gemini 2.5 Flash Image)Google · vision | $0.3 | $2.50 | $0.03 | · | 33K | 8K | ||
| Gemini 2.5 Flash LiteGoogle · reasoning · vision | $0.1 | $0.4 | $0.01 | $0.05 / $0.2 | 1.0M | 66K | ||
| Gemini 2.5 FlashGoogle · reasoning · vision | $0.3 | $2.50 | $0.03 | $0.15 / $1.25 | 1.0M | 66K | ||
| Gemini 2.5 ProGoogle · reasoning · vision | $1.25 | $10.0 | $0.125 | $0.625 / $5.00 | 1.0M | 66K | ||
| Gemini 2.5 Pro Preview 06-05Google · reasoning · vision | $1.25 | $10.0 | $0.125 | · | 1.0M | 66K | ||
| Gemma 3 4BGoogle · vision | $0.05 | $0.1 | · | · | 131K | 16K | ||
| Gemma 3 12BGoogle · vision | $0.05 | $0.15 | · | · | 131K | 16K | ||
| Gemma 3 27BGoogle · vision · 1 change | $0.08 | $0.35 | $0.04 | · | 131K | 118K | ||
| Gemma 2 27BGoogle | $0.65 | $0.65 | · | · | 8K | 2K | ||
| Muse Spark OpenMeta · reasoning · vision | $1.25 | $4.25 | $0.15 | · | 1.0M | 944K | ||
| Muse WonderMeta · reasoning · vision | $4.00 | $20.0 | $0.2 | · | 1.0M | 944K | ||
| Muse Spark 1.3 ContributorMeta · reasoning · vision | $0.1 | $0.2 | $0.002 | · | 1.0M | 944K | ||
| Muse Spark 1.3Meta · reasoning · vision | $1.25 | $4.25 | $0.15 | · | 1.0M | 944K | ||
| Muse Spark 1.2 ContributorMeta · reasoning · vision | $0.1 | $0.2 | $0.002 | · | 1.0M | 944K | ||
| Muse Glimmer 30BMeta · reasoning · vision · 6 changes | $0.3 | $1.20 | $0.04 | · | 131K | 16K | ||
| Muse Spark 1.2Meta · reasoning · vision | $1.25 | $4.25 | $0.15 | · | 1.0M | 944K | ||
| Muse Spark 1.1Meta · reasoning · vision | $1.25 | $4.25 | $0.15 | · | 1.0M | 944K | ||
| Llama Guard 4 12BMeta · vision | $0.18 | $0.18 | · | · | 164K | 16K | ||
| Llama 4 MaverickMeta · vision · 5 changes | $0.188 | $0.652 | $0.05 | · | 524K | 16K | ||
| Llama 4 ScoutMeta · vision · 2 changes | $0.1 | $0.3 | · | · | 1.3M | 16K | ||
| Llama 3.3 70B InstructMeta · 5 changes | $0.1 | $0.32 | · | · | 131K | 16K | ||
| Llama 3.2 1B InstructMeta | $0.027 | $0.201 | · | · | 60K | 54K | ||
| Llama 3.2 3B InstructMeta · 1 change | $0.051 | $0.335 | · | · | 131K | 72K | ||
| Llama 3.1 70B InstructMeta · 6 changes | $0.4 | $0.4 | · | · | 131K | 16K | ||
| Llama 3.1 8B InstructMeta | $0.05 | $0.08 | $0.025 | · | 131K | 118K | ||
| Phi 4Microsoft | $0.07 | $0.14 | · | · | 16K | 15K | ||
| GPT-6.1 Sol ProOpenAI · reasoning · vision | $2.00 | $10.0 | $0.1 | $1.00 / $5.00 | 1.1M | 128K | ||
| GPT-6.1 SolOpenAI · reasoning · vision | $2.00 | $10.0 | $0.1 | $1.00 / $5.00 | 1.1M | 128K | ||
| GPT-6 Luna ProOpenAI · reasoning · vision | $0.1 | $0.5 | $0.01 | $0.05 / $0.25 | 1.1M | 128K | ||
| GPT-6 LunaOpenAI · reasoning · vision | $0.1 | $0.5 | $0.01 | $0.05 / $0.25 | 1.1M | 128K | ||
| GPT-6 Sol ProOpenAI · reasoning · vision | $2.00 | $10.0 | $0.2 | $1.00 / $5.00 | 1.1M | 128K | ||
| GPT-6 SolOpenAI · reasoning · vision | $2.00 | $10.0 | $0.2 | $1.00 / $5.00 | 1.1M | 128K | ||
| GPT-6 AstraOpenAI · reasoning · vision | $10.0 | $50.0 | $1.00 | $5.00 / $25.0 | 1.1M | 128K | ||
| GPT-6 Astra ProOpenAI · reasoning · vision | $10.0 | $50.0 | $1.00 | $5.00 / $25.0 | 1.1M | 128K | ||
| GPT-5.6 Luna ProOpenAI · reasoning · vision | $0.2 | $1.20 | $0.02 | $0.1 / $0.6 | 1.1M | 128K | ||
| GPT-5.6 LunaOpenAI · reasoning · vision | $0.2 | $1.20 | $0.02 | $0.1 / $0.6 | 1.1M | 128K | ||
| GPT-5.6 Terra ProOpenAI · reasoning · vision | $2.00 | $12.0 | $0.2 | $1.00 / $6.00 | 1.1M | 128K | ||
| GPT-5.6 TerraOpenAI · reasoning · vision | $2.00 | $12.0 | $0.2 | $1.00 / $6.00 | 1.1M | 128K | ||
| GPT-5.6 Sol ProOpenAI · reasoning · vision · 3 changes | $4.00 | $20.0 | $0.4 | $2.00 / $10.0 | 1.1M | 128K | ||
| GPT-5.6 SolOpenAI · reasoning · vision · 1 change | $4.00 | $20.0 | $0.4 | $2.00 / $10.0 | 1.1M | 128K | ||
| GPT-5.5 ProOpenAI · reasoning · vision | $30.0 | $180 | · | $15.0 / $90.0 | 1.1M | 128K | ||
| GPT-5.5OpenAI · reasoning · vision | $5.00 | $30.0 | $0.5 | $2.50 / $15.0 | 1.1M | 128K | ||
| GPT-5.4 Image 2OpenAI · reasoning · vision | $8.00 | $15.0 | $2.00 | · | 272K | 128K | ||
| GPT-5.4 NanoOpenAI · reasoning · vision | $0.2 | $1.25 | $0.02 | $0.1 / $0.625 | 400K | 128K | ||
| GPT-5.4 MiniOpenAI · reasoning · vision | $0.75 | $4.50 | $0.075 | $0.375 / $2.25 | 400K | 128K | ||
| GPT-5.4 ProOpenAI · reasoning · vision | $30.0 | $180 | · | $15.0 / $90.0 | 1.1M | 128K | ||
| GPT-5.4OpenAI · reasoning · vision | $2.50 | $15.0 | $0.25 | $1.25 / $7.50 | 1.1M | 128K | ||
| GPT-5.3-CodexOpenAI · reasoning · vision | $1.75 | $14.0 | $0.175 | · | 400K | 128K | ||
| GPT AudioOpenAI | $2.50 | $10.0 | · | · | 128K | 16K | ||
| GPT Audio MiniOpenAI | $0.6 | $2.40 | · | · | 128K | 16K | ||
| GPT-5.2-CodexOpenAI · reasoning · vision | $1.75 | $14.0 | $0.175 | · | 400K | 128K | ||
| GPT-5.2 ChatOpenAI · vision | $1.75 | $14.0 | $0.175 | · | 128K | 32K | ||
| GPT-5.2 ProOpenAI · reasoning · vision | $21.0 | $168 | · | $10.5 / $84.0 | 400K | 128K | ||
| GPT-5.2OpenAI · reasoning · vision | $1.75 | $14.0 | $0.175 | $0.875 / $7.00 | 400K | 128K | ||
| GPT-5.1-Codex-MaxOpenAI · reasoning · vision | $1.25 | $10.0 | $0.125 | · | 400K | 128K | ||
| GPT-5.1OpenAI · reasoning · vision | $1.25 | $10.0 | $0.125 | $0.625 / $5.00 | 400K | 128K | ||
| GPT-5.1-CodexOpenAI · reasoning · vision | $1.25 | $10.0 | $0.13 | · | 400K | 128K | ||
| GPT-5.1-Codex-MiniOpenAI · reasoning · vision | $0.25 | $2.00 | $0.03 | · | 400K | 128K | ||
| gpt-oss-safeguard-20bOpenAI · reasoning | $0.075 | $0.3 | $0.037 | · | 131K | 66K | ||
| GPT-5 Image MiniOpenAI · reasoning · vision | $2.50 | $2.00 | $0.25 | · | 400K | 128K | ||
| GPT-5 ImageOpenAI · reasoning · vision | $10.0 | $10.0 | $1.25 | · | 400K | 128K | ||
| GPT-5 ProOpenAI · reasoning · vision | $15.0 | $120 | · | $7.50 / $60.0 | 400K | 128K | ||
| GPT-5OpenAI · reasoning · vision | $1.25 | $10.0 | $0.125 | $0.625 / $5.00 | 400K | 128K | ||
| GPT-5 MiniOpenAI · reasoning · vision | $0.25 | $2.00 | $0.025 | $0.125 / $1.00 | 400K | 128K | ||
| GPT-5 NanoOpenAI · reasoning · vision | $0.05 | $0.4 | $0.005 | $0.025 / $0.2 | 400K | 128K | ||
| gpt-oss-120bOpenAI · reasoning · 7 changes | $0.15 | $0.6 | $0.075 | $0.03 / $0.136 | 131K | 66K | ||
| gpt-oss-20bOpenAI · reasoning · 2 changes | $0.018 | $0.09 | $0.009 | $0.024 / $0.112 | 131K | 33K | ||
| o3 ProOpenAI · reasoning · vision | $20.0 | $80.0 | · | · | 200K | 100K | ||
| o4 Mini HighOpenAI · reasoning · vision | $1.10 | $4.40 | $0.275 | · | 200K | 100K | ||
| o3OpenAI · reasoning · vision | $2.00 | $8.00 | $0.5 | $1.00 / $4.00 | 200K | 100K | ||
| o4 MiniOpenAI · reasoning · vision | $1.10 | $4.40 | $0.275 | $0.55 / $2.20 | 200K | 100K | ||
| GPT-4.1OpenAI · vision | $2.00 | $8.00 | $0.5 | $1.00 / $4.00 | 1.0M | 33K | ||
| GPT-4.1 MiniOpenAI · vision | $0.4 | $1.60 | $0.1 | $0.2 / $0.8 | 1.0M | 33K | ||
| GPT-4.1 NanoOpenAI · vision | $0.1 | $0.4 | $0.025 | $0.05 / $0.2 | 1.0M | 33K | ||
| o1-proOpenAI · reasoning · vision | $150 | $600 | · | · | 200K | 100K | ||
| o3 Mini HighOpenAI · reasoning | $1.10 | $4.40 | $0.55 | · | 200K | 100K | ||
| o3 MiniOpenAI · reasoning | $1.10 | $4.40 | $0.55 | $0.55 / $2.20 | 200K | 100K | ||
| o1OpenAI · reasoning · vision | $15.0 | $60.0 | $7.50 | · | 200K | 100K | ||
| GPT-4o (2024-11-20)OpenAI · vision | $2.50 | $10.0 | $1.25 | · | 128K | 16K | ||
| GPT-4o (2024-08-06)OpenAI · vision | $2.50 | $10.0 | $1.25 | · | 128K | 16K | ||
| GPT-4o-miniOpenAI · vision | $0.15 | $0.6 | $0.075 | $0.075 / $0.3 | 128K | 16K | ||
| GPT-4o-mini (2024-07-18)OpenAI · vision | $0.15 | $0.6 | $0.075 | · | 128K | 16K | ||
| GPT-4oOpenAI · vision | $2.50 | $10.0 | $1.25 | $1.25 / $5.00 | 128K | 16K | ||
| GPT-4o (2024-05-13)OpenAI · vision | $5.00 | $15.0 | · | · | 128K | 4K | ||
| GPT-4 TurboOpenAI · vision | $10.0 | $30.0 | · | $5.00 / $15.0 | 128K | 4K | ||
| GPT-3.5 Turbo (older v0613)OpenAI | $1.00 | $2.00 | · | · | 4K | 4K | ||
| GPT-3.5 Turbo InstructOpenAI | $1.50 | $2.00 | · | · | 4K | 4K | ||
| GPT-3.5 Turbo 16kOpenAI | $3.00 | $4.00 | · | · | 16K | 4K | ||
| GPT-3.5 TurboOpenAI | $0.5 | $1.50 | · | $0.25 / $0.75 | 16K | 4K | ||
| GPT-4OpenAI | $30.0 | $60.0 | · | · | 8K | 4K | ||
| Grok 4.7xAI · reasoning · vision · 1 change | $2.00 | $6.00 | $0.5 | · | 500K | 450K | ||
| Grok 4.6xAI · reasoning · vision | $2.00 | $6.00 | $0.5 | · | 500K | 450K | ||
| Grok 4.5xAI · reasoning · vision | $2.00 | $6.00 | $0.3 | · | 500K | 450K | ||
| Grok Build 0.1xAI · reasoning · vision | $1.00 | $2.00 | $0.2 | · | 256K | 230K | ||
| Grok 4.3xAI · reasoning · vision | $1.25 | $2.50 | $0.2 | $1.00 / $2.00 | 1M | 900K | ||
| Grok 4.20 Multi-AgentxAI · reasoning · vision | $1.25 | $2.50 | $0.2 | · | 2M | 1.8M | ||
| Grok 4.20xAI · reasoning · vision | $1.25 | $2.50 | $0.2 | · | 2M | 1.8M |
No model matches. Clear a filter.
Price history
Compare up to 10 models over any date range. Pick the price you pay for, tick the models, drag the dates.
Move over the chart to read prices on a date.
5 of 10 picked ·
OpenAI
Anthropic
Microsoft
Meta
xAI
USD per million tokens, list price from the OpenRouter model list. A line holds until the price changes; the log axis keeps cheap and expensive models on one chart. History starts Sep 9, 2026, last checked Oct 19, 2026. Series JSON · Embed this chart.
Price change log
| Date | Model | Change |
|---|---|---|
| Llama 3.2 3B Instruct Meta | input: $0.05 → $0.051 per M | |
| Llama 3.2 3B Instruct Meta | output: $0.33 → $0.335 per M | |
| Llama 3.2 3B Instruct Meta | max output: 118K → 72K | |
| Gemini 4 Argon Google | max output: 66K → 944K | |
| Gemma 4 26B A4B Google | input: $0.068 → $0.09 per M | |
| Gemma 4 26B A4B Google | output: $0.225 → $0.3 per M | |
| Gemma 4 26B A4B Google | cached input: $0.037 → $0.05 per M | |
| Llama 4 Maverick Meta | context: 1.0M → 524K | |
| WizardLM-2 8x22B Microsoft | delisted (was $0.62 in / $0.62 out) | |
| Gemma 3 27B Google | output: $0.45 → $0.35 per M | |
| Gemma 4 26B A4B Google | input: $0.09 → $0.068 per M | |
| Gemma 4 26B A4B Google | output: $0.3 → $0.225 per M | |
| Gemma 4 26B A4B Google | cached input: $0.05 → $0.037 per M | |
| Gemini 3.8 Flash Google | input: $0.75 → $0.825 per M | |
| Gemini 3.8 Flash Google | output: $3.75 → $4.13 per M | |
| Gemini 3.8 Flash Google | cached input: $0.075 → $0.083 per M | |
| Gemma 4 26B A4B Google | input: $0.09 → $0.076 per M | |
| Gemma 4 26B A4B Google | output: $0.3 → $0.255 per M | |
| Gemma 4 26B A4B Google | cached input: $0.05 → $0.043 per M | |
| gpt-oss-120b OpenAI | input: $0.15 → $0.037 per M | |
| gpt-oss-120b OpenAI | output: $0.6 → $0.17 per M | |
| gpt-oss-120b OpenAI | cached input: $0.075 → per M | |
| gpt-oss-120b OpenAI | max output: 66K → 118K | |
| Gemma 4 26B A4B Google | input: $0.076 → $0.09 per M | |
| Gemma 4 26B A4B Google | output: $0.255 → $0.3 per M | |
| Gemma 4 26B A4B Google | cached input: $0.043 → $0.05 per M | |
| gpt-oss-120b OpenAI | input: $0.037 → $0.15 per M | |
| gpt-oss-120b OpenAI | output: $0.17 → $0.6 per M | |
| gpt-oss-120b OpenAI | cached input: → $0.075 per M | |
| gpt-oss-120b OpenAI | max output: 118K → 66K | |
| Llama 3.3 70B Instruct Meta | input: $0.59 → $0.1 per M | |
| Llama 3.3 70B Instruct Meta | output: $0.79 → $0.32 per M | |
| Llama 3.3 70B Instruct Meta | cached input: $0.295 → per M | |
| Llama 3.3 70B Instruct Meta | max output: 33K → 16K | |
| Llama 4 Scout Meta | input: $0.18 → $0.1 per M | |
| Llama 4 Scout Meta | output: $0.59 → $0.3 per M | |
| Llama 4 Scout Meta | max output: 118K → 16K | |
| GPT-5.6 Sol Pro OpenAI | input: $2.00 → $4.00 per M | |
| GPT-5.6 Sol Pro OpenAI | output: $10.0 → $20.0 per M | |
| GPT-5.6 Sol Pro OpenAI | cached input: $0.2 → $0.4 per M | |
| GPT-5.6 Sol OpenAI | input: $2.00 → $4.00 per M | |
| GPT-5.6 Sol OpenAI | output: $10.0 → $20.0 per M | |
| GPT-5.6 Sol OpenAI | cached input: $0.2 → $0.4 per M | |
| Gemma 4 26B A4B Google | input: $0.09 → $0.076 per M | |
| Gemma 4 26B A4B Google | output: $0.3 → $0.255 per M | |
| Gemma 4 26B A4B Google | cached input: $0.05 → $0.043 per M | |
| Gemma 4 31B Google | input: $0.09 → $0.2 per M | |
| Gemma 4 31B Google | output: $0.34 → $0.4 per M | |
| Gemma 4 31B Google | cached input: $0.05 → per M | |
| Gemma 4 26B A4B Google | input: $0.076 → $0.09 per M | |
| Gemma 4 26B A4B Google | output: $0.255 → $0.3 per M | |
| Gemma 4 26B A4B Google | cached input: $0.043 → $0.05 per M | |
| gpt-oss-120b OpenAI | input: $0.037 → $0.15 per M | |
| gpt-oss-120b OpenAI | output: $0.17 → $0.6 per M | |
| gpt-oss-120b OpenAI | cached input: → $0.075 per M | |
| gpt-oss-120b OpenAI | max output: 118K → 66K | |
| Muse Spark Open Meta | listed at $1.25 in / $4.25 out | |
| Claude Sonnet 4 Anthropic | delisted (was $3 in / $15 out) | |
| Gemma 4 26B A4B Google | input: $0.076 → $0.09 per M | |
| Gemma 4 26B A4B Google | output: $0.255 → $0.3 per M | |
| Gemma 4 26B A4B Google | cached input: $0.043 → $0.05 per M | |
| gpt-oss-120b OpenAI | input: $0.15 → $0.037 per M | |
| gpt-oss-120b OpenAI | output: $0.6 → $0.17 per M | |
| gpt-oss-120b OpenAI | cached input: $0.075 → per M | |
| gpt-oss-120b OpenAI | max output: 66K → 118K | |
| Muse Wonder Meta | listed at $4 in / $20 out | |
| Gemma 4 26B A4B Google | input: $0.09 → $0.076 per M | |
| Gemma 4 26B A4B Google | output: $0.3 → $0.255 per M | |
| Gemma 4 26B A4B Google | cached input: $0.05 → $0.043 per M | |
| gpt-oss-120b OpenAI | input: $0.037 → $0.15 per M | |
| gpt-oss-120b OpenAI | output: $0.17 → $0.6 per M | |
| gpt-oss-120b OpenAI | cached input: → $0.075 per M | |
| gpt-oss-120b OpenAI | max output: 118K → 66K | |
| Llama 3.3 70B Instruct Meta | input: $0.22 → $0.59 per M | |
| Llama 3.3 70B Instruct Meta | output: $0.5 → $0.79 per M | |
| Llama 3.3 70B Instruct Meta | cached input: $0.11 → $0.295 per M | |
| Llama 3.3 70B Instruct Meta | max output: 16K → 33K | |
| Gemma 4 26B A4B Google | input: $0.076 → $0.09 per M | |
| Gemma 4 26B A4B Google | output: $0.255 → $0.3 per M | |
| Gemma 4 26B A4B Google | cached input: $0.043 → $0.05 per M |
By company
OpenAI 64 priced models. Cheapest output $0.09/M (gpt-oss-20b); top of the range $600/M (o1-pro). Largest context 1.1M. Model docs and cards · Developer posts
Google 30 priced models. Cheapest output $0.0/M (Lyria 3 Pro Preview); top of the range $12.0/M (Nano Banana Pro (Gemini 3 Pro Image Preview)). Largest context 1.0M. Model docs and cards · Developer posts
Anthropic 15 priced models. Cheapest output $0.5/M (Claude Haiku 5.5); top of the range $75.0/M (Claude Opus 4.1). Largest context 1M. Model docs and cards · Developer posts
Microsoft 1 priced models. Cheapest output $0.14/M (Phi 4); top of the range $0.14/M (Phi 4). Largest context 16K. Model docs and cards · Developer posts
Meta 16 priced models. Cheapest output $0.08/M (Llama 3.1 8B Instruct); top of the range $20.0/M (Muse Wonder). Largest context 1.3M. Model docs and cards · Developer posts
xAI 7 priced models. Cheapest output $2.00/M (Grok Build 0.1); top of the range $6.00/M (Grok 4.5). Largest context 2M. Model docs and cards · Developer posts
Questions people ask
- Where do these prices come from?
- From the public OpenRouter model list, refreshed on every Subvolts pipeline run (three times a day). OpenRouter passes through each provider's list price, so the numbers match the labs' own pricing pages for the same model. Prices are USD per million tokens, before any volume discount, batch discount or committed-use deal.
- What do "batch" prices mean?
- Most labs sell a batch tier: send a file of requests, get answers within hours instead of seconds, pay about half. The batch column shows that input and output price where the provider lists one.
- What does "cached input" mean?
- Most providers charge less for input tokens they have already processed in a recent request (prompt caching). If your app repeats a long system prompt or document, that part is billed at the cached rate. The calculator lets you set the share of input you expect to hit the cache.
- Why is a model missing?
- The table shows models the six labs sell through an API and that OpenRouter lists. Consumer-only models (inside ChatGPT, Gemini app, Copilot) and models sold only through a private cloud contract are not priced per token and do not appear. Open-weight models show the cheapest hosted price OpenRouter lists.
- How do I read the change log?
- Every day the pipeline compares each model's price, context window and output limit with the previous day. Any difference becomes a dated row. "Listed" means the model appeared for the first time; "delisted" means it dropped off the list.
- Can I use this data?
- Yes. Download the CSV or JSON, or embed the live table with the widget link in the sidebar. Credit Subvolts with a link to this page.
Methodology: prices are list prices per million tokens in USD from the public OpenRouter model list, which mirrors each lab's own pricing page for the same model id. Consumer plans are not priced per token and live in the plan comparison (coming). Report an error via contact. Editorial standards.