LLM pricing, by provider
What the model APIs CostCompass tracks charge per million tokens, side by side: input, output, and the cache rates that decide most of an agent's bill. These are the list prices CostCompass uses to estimate spend, and each one is compared with the provider's own pricing page once a week. The log at the bottom lists every price change that check found.
Prices per 1M tokens, in US dollars, as each provider lists them. Last change .
Anthropic (Claude)
| Model | Input | Output |
|---|---|---|
| Claude Fable 5.1 | 10.00 | 50.00 |
| Claude Fable 5.1 · cache read | 0.25 | — |
| Claude Fable 5.1 · cache write | 12.50 | — |
| Claude Fable 5.1 · cache write 1h | 20.00 | — |
| Claude Fable 5.2 | 8.00 | 40.00 |
| Claude Fable 5.2 · cache read | 0.20 | — |
| Claude Fable 5.2 · cache write | 10.00 | — |
| Claude Fable 5.2 · cache write 1h | 16.00 | — |
| Claude Haiku 5.5 | 0.10 | 0.50 |
| Claude Haiku 5.5 · cache read | 0.01 | — |
| Claude Haiku 5.5 · cache write | 0.125 | — |
| Claude Haiku 5.5 · cache write 1h | 0.20 | — |
| Claude Mythos 5 | 10.00 | 50.00 |
| Claude Mythos 5 · cache read | 1.00 | — |
| Claude Mythos 5 · cache write | 12.50 | — |
| Claude Mythos 5 · cache write 1h | 20.00 | — |
| Claude Mythos 5.1 | 10.00 | 50.00 |
| Claude Mythos 5.1 · cache read | 0.25 | — |
| Claude Mythos 5.1 · cache write | 12.50 | — |
| Claude Mythos 5.1 · cache write 1h | 20.00 | — |
| Claude Mythos 5.2 | 8.00 | 40.00 |
| Claude Mythos 5.2 · cache read | 0.20 | — |
| Claude Mythos 5.2 · cache write | 10.00 | — |
| Claude Mythos 5.2 · cache write 1h | 16.00 | — |
| Claude Opus 5.5 | 4.00 | 20.00 |
| Claude Opus 5.5 · cache read | 0.20 | — |
| Claude Opus 5.5 · cache write | 5.00 | — |
| Claude Opus 5.5 · cache write 1h | 8.00 | — |
| Claude Sonnet 5.5 | 2.00 | 10.00 |
| Claude Sonnet 5.5 · cache read | 0.10 | — |
| Claude Sonnet 5.5 · cache write | 2.50 | — |
| Claude Sonnet 5.5 · cache write 1h | 4.00 | — |
DeepSeek
| Model | Input | Output |
|---|---|---|
| DeepSeek Flash | 0.30 | 1.20 |
| DeepSeek Flash · cache read | 0.006 | — |
| DeepSeek Flash · cache read · off peak | 0.003 | — |
| DeepSeek Flash · off peak | 0.15 | 0.60 |
| DeepSeek V4 Pro | 1.32 | 3.96 |
| DeepSeek V4 Pro · cache read | 0.044 | — |
| DeepSeek V4 Pro · cache read · off peak | 0.022 | — |
| DeepSeek V4 Pro · off peak | 0.66 | 1.98 |
Rates are time-of-day. Peak is 01:00–04:00 and 06:00–10:00 UTC, Monday to Friday; every other hour — and all weekend — is off-peak at half the peak rate. Input is the cache-miss rate. Search and Fetch are priced per successful request, at any hour.
| Model | Input | Output |
|---|---|---|
| Gemini 2.5 Flash | 0.30 | 2.50 |
| Gemini 2.5 Flash · cached input | 0.03 | — |
| Gemini 2.5 Flash · thinking output | — | 2.50 |
| Gemini 2.5 Flash Image | 0.30 | — |
| Gemini 2.5 Flash Image · image output | — | 30.00 |
| Gemini 2.5 Flash Lite | 0.10 | 0.40 |
| Gemini 2.5 Flash Lite · cached input | 0.01 | — |
| Gemini 2.5 Pro | 1.25 | 10.00 |
| Gemini 2.5 Pro · cached input | 0.125 | — |
| Gemini 2.5 Pro · thinking output | — | 10.00 |
| Gemini 3 Pro Image | 2.00 | 12.00 |
| Gemini 3 Pro Image · image output | — | 120.00 |
| Gemini 3.1 Flash Image | 0.50 | 3.00 |
| Gemini 3.1 Flash Image · image output | — | 60.00 |
| Gemini 3.1 Flash Lite | 0.25 | 1.50 |
| Gemini 3.1 Flash Lite · cached input | 0.025 | — |
| Gemini 3.1 Flash Lite · thinking output | — | 1.50 |
| Gemini 3.1 Flash Lite Image | 0.25 | 1.50 |
| Gemini 3.1 Flash Lite Image · image output | — | 30.00 |
| Gemini 3.1 Pro Preview | 2.00 | 12.00 |
| Gemini 3.1 Pro Preview · cached input | 0.20 | — |
| Gemini 3.1 Pro Preview · thinking output | — | 12.00 |
| Gemini 3.5 Flash | 1.50 | 9.00 |
| Gemini 3.5 Flash · cached input | 0.15 | — |
| Gemini 3.5 Flash · thinking output | — | 9.00 |
| Gemini 3.5 Flash Lite | 0.30 | 2.50 |
| Gemini 3.5 Flash Lite · cached input | 0.03 | — |
| Gemini 3.5 Flash Lite · thinking output | — | 2.50 |
| Gemini 3.6 Flashuntil , then 1.50 / 7.50 | 0.75 | 3.75 |
| Gemini 3.6 Flash · cached inputuntil , then 0.15 / — | 0.075 | — |
| Gemini 3.6 Flash · thinking outputuntil , then — / 7.50 | — | 3.75 |
| Gemini 3.7 Flashuntil , then 1.50 / 7.50 | 0.75 | 3.75 |
| Gemini 3.7 Flash · cached inputuntil , then 0.15 / — | 0.075 | — |
| Gemini 3.7 Flash · thinking outputuntil , then — / 7.50 | — | 3.75 |
| Gemini 3.8 Flashuntil , then 1.50 / 7.50 | 0.75 | 3.75 |
| Gemini 3.8 Flash · cached inputuntil , then 0.15 / — | 0.075 | — |
| Gemini 3.8 Flash · thinking outputuntil , then — / 7.50 | — | 3.75 |
| Gemini 3.8 Flash Lite Ttsuntil , then 1.00 / — | 0.50 | — |
| Gemini 3.8 Flash Lite Tts · audio outputuntil , then — / 12.00 | — | 6.00 |
| Gemini 3.8 Flash Ttsuntil , then 1.00 / — | 0.50 | — |
| Gemini 3.8 Flash Tts · audio outputuntil , then — / 18.00 | — | 9.00 |
| Gemini 4 Argonuntil , then 4.00 / 20.00 | 2.00 | 10.00 |
| Gemini 4 Argon · cached inputuntil , then 0.20 / — | 0.10 | — |
| Gemini 4 Argon · thinking outputuntil , then — / 20.00 | — | 10.00 |
OpenAI
| Model | Input | Output |
|---|---|---|
| chat-latest | 5.00 | 30.00 |
| chat-latest · cache read | 0.50 | — |
| GPT-4.1 | 2.00 | 8.00 |
| GPT-4.1 · cache read | 0.50 | — |
| GPT-4.1 mini | 0.40 | 1.60 |
| GPT-4.1 mini · cache read | 0.10 | — |
| GPT-4o | 2.50 | 10.00 |
| GPT-4o · cache read | 1.25 | — |
| GPT-4o mini | 0.15 | 0.60 |
| GPT-4o mini · cache read | 0.075 | — |
| GPT-5 | 1.25 | 10.00 |
| GPT-5 · cache read | 0.125 | — |
| GPT-5 mini | 0.25 | 2.00 |
| GPT-5 mini · cache read | 0.025 | — |
| GPT-5 nano | 0.05 | 0.40 |
| GPT-5 nano · cache read | 0.005 | — |
| GPT-5 Pro | 15.00 | 120.00 |
| GPT-5 Search Api | 1.25 | 10.00 |
| GPT-5 Search Api · cache read | 0.125 | — |
| GPT-5.1 | 1.25 | 10.00 |
| GPT-5.1 · cache read | 0.125 | — |
| GPT-5.2 | 1.75 | 14.00 |
| GPT-5.2 · cache read | 0.175 | — |
| GPT-5.2 Pro | 21.00 | 168.00 |
| GPT-5.3 Codex | 1.75 | 14.00 |
| GPT-5.3 Codex · cache read | 0.175 | — |
| GPT-5.4 | 2.50 | 15.00 |
| GPT-5.4 · cache read | 0.25 | — |
| GPT-5.4 mini | 0.75 | 4.50 |
| GPT-5.4 mini · cache read | 0.075 | — |
| GPT-5.4 nano | 0.20 | 1.25 |
| GPT-5.4 nano · cache read | 0.02 | — |
| GPT-5.4 Pro | 30.00 | 180.00 |
| GPT-5.5 | 5.00 | 30.00 |
| GPT-5.5 · cache read | 0.50 | — |
| GPT-5.5 Cyber Preview | 12.50 | 75.00 |
| GPT-5.5 Cyber Preview · cache read | 1.25 | — |
| GPT-5.5 Pro | 30.00 | 180.00 |
| GPT-5.6 Cyber | 12.50 | 75.00 |
| GPT-5.6 Cyber · cache read | 1.25 | — |
| GPT-5.6 Cyber · cache write | 15.625 | — |
| GPT-5.6 Luna | 0.20 | 1.20 |
| GPT-5.6 Luna · cache read | 0.02 | — |
| GPT-5.6 Luna · cache write | 0.25 | — |
| GPT-5.6 Sol | 4.00 | 20.00 |
| GPT-5.6 Sol · cache read | 0.40 | — |
| GPT-5.6 Sol · cache write | 5.00 | — |
| GPT-5.6 Terra | 2.00 | 12.00 |
| GPT-5.6 Terra · cache read | 0.20 | — |
| GPT-5.6 Terra · cache write | 2.50 | — |
| GPT-6 Astra | 10.00 | 50.00 |
| GPT-6 Astra · cache read | 1.00 | — |
| GPT-6 Astra · cache write | 12.50 | — |
| GPT-6 Astra Law | 10.00 | 50.00 |
| GPT-6 Astra Law · cache read | 1.00 | — |
| GPT-6 Cyber | 20.00 | 100.00 |
| GPT-6 Cyber · cache read | 2.00 | — |
| GPT-6 Cyber · cache write | 25.00 | — |
| GPT-6 Luna | 0.10 | 0.50 |
| GPT-6 Luna · cache read | 0.01 | — |
| GPT-6 Luna · cache write | 0.125 | — |
| GPT-6 Rosalind | 15.00 | 75.00 |
| GPT-6 Rosalind · cache read | 1.50 | — |
| GPT-6 Rosalind · cache write | 18.75 | — |
| GPT-6 Sol | 2.00 | 10.00 |
| GPT-6 Sol · cache read | 0.20 | — |
| GPT-6 Sol · cache write | 2.50 | — |
| GPT-6.1 Sol | 2.00 | 10.00 |
| GPT-6.1 Sol · cache read | 0.10 | — |
| GPT-6.1 Sol · cache write | 2.50 | — |
| GPT-6.1 Terra | 1.00 | 5.00 |
| GPT-6.1 Terra · cache read | 0.05 | — |
| GPT-6.1 Terra · cache write | 1.25 | — |
| GPT-rosalind Discovery | 5.00 | 25.00 |
| GPT-rosalind Discovery · cache read | 0.50 | — |
| GPT-rosalind Research | 5.00 | 25.00 |
| GPT-rosalind Research · cache read | 0.50 | — |
| o3 | 2.00 | 8.00 |
| o3 · cache read | 0.50 | — |
| o3-pro | 20.00 | 80.00 |
| text-embedding-3-large | 0.13 | — |
| text-embedding-3-small | 0.02 | — |
| text-embedding-ada-002 | 0.10 | — |
Standard short-context text rates; cache read is the cached-input rate and cache write is the short-context 30-minute cache-write rate, on the models billed for one. Long-context prompts, processing tiers, and regional processing may have different rates. Spend uses OpenAI's billed amounts.
xAI (Grok)
| Model | Input | Output |
|---|---|---|
| Grok 4.20 0309 Non Reasoning | 1.25 | 2.50 |
| Grok 4.20 0309 Non Reasoning · cache read | 0.20 | — |
| Grok 4.20 0309 Reasoning | 1.25 | 2.50 |
| Grok 4.20 0309 Reasoning · cache read | 0.20 | — |
| Grok 4.20 Multi Agent 0309 | 1.25 | 2.50 |
| Grok 4.20 Multi Agent 0309 · cache read | 0.20 | — |
| Grok 4.3 | 1.25 | 2.50 |
| Grok 4.3 · cache read | 0.20 | — |
| Grok 4.5 | 2.00 | 6.00 |
| Grok 4.5 · cache read | 0.30 | — |
| Grok 4.6 | 2.00 | 6.00 |
| Grok 4.6 · cache read | 0.50 | — |
| Grok 4.7 | 2.00 | 6.00 |
| Grok 4.7 · cache read | 0.50 | — |
| Grok Build 0.1 | 1.00 | 2.00 |
| Grok Build 0.1 · cache read | 0.20 | — |
Rates shown are for prompts up to 200k tokens. Longer prompts are billed at twice these rates.
Price changes
Every price change the weekly check recorded, newest first, since .
| Date | Provider | Model | Change | Before | After |
|---|---|---|---|---|---|
| OpenAI | GPT-5.5 Cyber Preview | New model | — | 12.50 / 75.00 | |
| Anthropic (Claude) | Claude Mythos 5.2 | New model | — | 8.00 / 40.00 | |
| Anthropic (Claude) | Claude Fable 5.2 | New model | — | 8.00 / 40.00 | |
| OpenAI | GPT-6 Astra Law | New model | — | 10.00 / 50.00 | |
| OpenAI | GPT-6 Cyber | New model | — | 20.00 / 100.00 | |
| OpenAI | GPT-6 Rosalind | New model | — | 15.00 / 75.00 | |
| OpenAI | GPT-rosalind Discovery | New model | — | 5.00 / 25.00 | |
| OpenAI | GPT-rosalind Research | New model | — | 5.00 / 25.00 | |
| OpenAI | GPT-6.1 Terra | New model | — | 1.00 / 5.00 | |
| Anthropic (Claude) | Claude Sonnet 4.5 | Retired | 3.00 / 15.00 | — |
Reading these prices
A list price is where your bill starts, not what it ends at. Two models at the same price per token can cost very different amounts for the same text, because each one counts tokens with its own tokenizer. The token cost calculator shows that difference on your own text.
Cache rows matter more than they look. A cache read usually costs a fraction of the input price, so a long conversation that reuses its context can bill well below the headline rate. A cache write can cost more than plain input. Rows marked off peak apply only during the hours the provider names.
Discounts, batch rates, regional pricing and credits are not in these tables. Your provider's invoice is the final word on what you were charged.