Short answer
APIVAI sells 23 models through one API key as of 2026-11-18: 13 Claude models from Anthropic and 10 GPT models from OpenAI. Claude models cost up to 63% less than Anthropic's list prices and GPT models up to 86% less than OpenAI's. Prices below are per million tokens; each model's page has its cost per request, code examples and FAQ.
Which model should I choose?
Start with Sonnet for most API work and move one tier up or down only when your own tests call for it. Anthropic positions its four Claude tiers this way in its models overview: Haiku is the fastest and cheapest, for high-volume classification, extraction and routing; Sonnet balances speed and intelligence for everyday coding, analysis, writing and tool use; Opus is for long-running agentic coding and hard reasoning; Fable is the most capable and most expensive tier and pays off only when Opus falls short. On the GPT side, Luna models are the cheapest on the list and suit bulk summarization and routing, Astra is the top tier, and Sol and Terra sit in between on price. One key works for every model here, so you can pick a model per task instead of committing to one. Test your prompts on two neighbouring tiers: if the cheaper one passes, keep it; if it needs many retries, the stronger model is often cheaper overall. Per-task recommendations are in Sonnet vs Opus vs GPT: API cost compared, and the cost calculator prices your own workload on every model.
Claude models (Anthropic)
| Model | Model ID | Input / output | Official price | Saving | Context |
|---|---|---|---|---|---|
| Claude Fable 5.2 | claude-fable-5-2 | $3.18 / $15.90 | $8.00 / $40.00 | 60% | 1M |
| Claude Fable 5.1 | claude-fable-5-1 | $3.97 / $19.87 | $10.00 / $50.00 | 60% | 1M |
| Claude Fable 5 | claude-fable-5 | $3.97 / $19.87 | $10.00 / $50.00 | 60% | 1M |
| Claude Opus 5.5 | claude-opus-5-5 | $1.52 / $7.57 | $4.00 / $20.00 | 62% | 1M |
| Claude Opus 5 | claude-opus-5 | $1.89 / $9.47 | $5.00 / $25.00 | 62% | 1M |
| Claude Opus 4.8 | claude-opus-4-8 | $1.89 / $9.47 | $5.00 / $25.00 | 62% | 1M |
| Claude Opus 4.7 | claude-opus-4-7 | $1.89 / $9.47 | $5.00 / $25.00 | 62% | 1M |
| Claude Opus 4.6 | claude-opus-4-6 | $1.89 / $9.47 | $5.00 / $25.00 | 62% | 1M |
| Claude Sonnet 5.5 | claude-sonnet-5-5 | $0.75 / $3.79 | $2.00 / $10.00 | 63% | 1M |
| Claude Sonnet 5 | claude-sonnet-5 | $0.75 / $3.79 | $2.00 / $10.00 | 63% | 1M |
| Claude Sonnet 4.6 | claude-sonnet-4-6 | $1.14 / $5.68 | $3.00 / $15.00 | 62% | 1M |
| Claude Haiku 5.5 | claude-haiku-5-5 | $0.19 / $0.94 | $0.10 / $0.50 | — | 1M |
| Claude Haiku 4.5 | claude-haiku-4-5 | $0.38 / $1.89 | $1.00 / $5.00 | 62% | 200K |
GPT models (OpenAI)
| Model | Model ID | Input / output | Official price | Saving | Context |
|---|---|---|---|---|---|
| GPT-6.2 Astra | gpt-6.2-astra | $1.22 / $6.08 | $8.00 / $40.00 | 85% | 1M |
| GPT-6.2 Sol | gpt-6.2-sol | $0.30 / $1.52 | $2.00 / $10.00 | 85% | 1M |
| GPT-6.1 Sol | gpt-6.1-sol | $0.30 / $1.52 | $2.00 / $10.00 | 85% | 1M |
| GPT-6.1 Terra | gpt-6.1-terra | $0.15 / $0.77 | $1.00 / $5.00 | 85% | 1M |
| GPT-6 Astra | gpt-6-astra | $1.52 / $7.60 | $10.00 / $50.00 | 85% | 1M |
| GPT-6 Sol | gpt-6-sol | $0.30 / $1.52 | $2.00 / $10.00 | 85% | 1M |
| GPT-6 Luna | gpt-6-luna | $0.0144 / $0.0768 | $0.10 / $0.50 | 86% | 1M |
| GPT-5.6 Sol | gpt-5.6-sol | $0.61 / $3.65 | $4.00 / $20.00 | 85% | 1M |
| GPT-5.6 Terra | gpt-5.6-terra | $0.30 / $1.82 | $2.00 / $12.00 | 85% | 1M |
| GPT-5.6 Luna | gpt-5.6-luna | $0.0304 / $0.18 | $0.20 / $1.20 | 85% | 1M |
Prices are checked against the price list every 6 hours. Claude models work in both the Anthropic Messages and the OpenAI format; GPT models use the OpenAI format. See pricing for cache prices and the API documentation for setup.