Live pricing for 23+ models across Anthropic, OpenAI, Google, Meta, Mistral, DeepSeek, and xAI. Sortable, filterable, with daily historical snapshots. Use it to compare costs, spot price changes, or just stop having to Google "claude sonnet price" every time you're sizing a deployment.
| Model | Provider | Tier ↑ | Input / M | Output / M | Out/In ratio | Context | Cached in / M |
|---|---|---|---|---|---|---|---|
| Claude Opus 4.6 | Anthropic | Frontier | $5.00 | $25.00 | 5.0× | 200K | $0.50 |
| GPT-5.2 Pro | OpenAI | Frontier | $21.00 | $168.00 | 8.0× | 400K | — |
| Claude Sonnet 4.6 | Anthropic | Premium | $3.00 | $15.00 | 5.0× | 200K | $0.30 |
| GPT-5.2 | OpenAI | Premium | $1.75 | $14.00 | 8.0× | 400K | $0.17 |
| GPT-4.1 | OpenAI | Premium | $2.00 | $8.00 | 4.0× | 1M | $0.50 |
| Gemini 2.5 Pro | Premium | $1.25 | $10.00 | 8.0× | 1M | $0.13 | |
| Mistral Large 3 | Mistral | Premium | $2.00 | $6.00 | 3.0× | 256K | — |
| Grok-3 | xAI | Premium | $3.00 | $15.00 | 5.0× | 1M | — |
| Claude Haiku 4.5 | Anthropic | Mid | $0.25 | $1.25 | 5.0× | 200K | $0.03 |
| GPT-4.1 Mini | OpenAI | Mid | $0.40 | $1.60 | 4.0× | 1M | $0.10 |
| GPT-5 Mini | OpenAI | Mid | $0.25 | $2.00 | 8.0× | 128K | — |
| Gemini 2.5 Flash | Mid | $0.30 | $2.50 | 8.3× | 1M | $0.03 | |
| Mistral Medium 3 | Mistral | Mid | $0.40 | $2.00 | 5.0× | 128K | — |
| DeepSeek-R1 | DeepSeek | Mid | $0.55 | $2.19 | 4.0× | 128K | $0.06 |
| GPT-5 Nano | OpenAI | Budget | $0.05 | $0.40 | 8.0× | 128K | — |
| Gemini 2.0 Flash | Budget | $0.10 | $0.40 | 4.0× | 1M | $0.01 | |
| Llama 4 Scout | Meta | Budget | $0.15 | $0.50 | 3.3× | 10M | — |
| Llama 4 Maverick | Meta | Budget | $0.22 | $0.85 | 3.9× | 1M | — |
| Llama 3.3 70B | Meta | Budget | $0.12 | $0.30 | 2.5× | 128K | — |
| Mistral Small 3.1 | Mistral | Budget | $0.10 | $0.30 | 3.0× | 128K | — |
| DeepSeek-V3 | DeepSeek | Budget | $0.14 | $0.28 | 2.0× | 128K | $0.01 |
| DeepSeek-V3.2 | DeepSeek | Budget | $0.28 | $0.42 | 1.5× | 128K | $0.03 |
| Grok-3 Mini | xAI | Budget | $0.30 | $0.50 | 1.7× | 1M | — |
Snapshotting begins on first deploy of the daily Cloud Function.
Every 24 hours we capture every model's pricing into a Firestore snapshot. Once 30+ days of history exist, this chart will show input + output price trends and surface cuts/raises across providers.
The JSON API at /api/llm-prices.json already exposes current prices for anyone who wants to track their own.
We notify subscribers when a major provider raises or cuts prices (typically a few times per quarter). Also get the AI Hardware Buyer's Guide 2026 — when API prices push you toward self-hosting, this is the playbook.
→ Want to know when self-hosting pays off vs paying these prices? Use the Self-Host vs API Calculator.
Pricing data is maintained by hand from each provider's official pricing page. When prices
change, the source-of-truth file src/data/llmModels.ts is updated and the next
deploy reflects new numbers. A scheduled Cloud Function takes a daily snapshot to Firestore,
building a time series. CC BY 4.0 — cite back to https://www.kunalganglani.com/llm-prices.
The JSON API at /api/llm-prices.json is regenerated at every build. Cache
friendly (Cache-Control: public, max-age=3600), CORS-open.
Notice a wrong price? Email [email protected] with the source URL.