AI Rate CardClaude Sonnet 5 vs GPT-5.6 Luna: API price and specs
GPT-5.6 Luna is 90% cheaper than Claude Sonnet 5 on input price.
GPT-5.6 Luna is 88% cheaper than Claude Sonnet 5 on output price.
GPT-5.6 Luna is 89% cheaper than Claude Sonnet 5 on a 1M-in / 200k-out workload.
GPT-5.6 Luna has a 5% larger context than Claude Sonnet 5.
| Field | Claude Sonnet 5 | GPT-5.6 Luna |
|---|---|---|
| Lowest input, per 1M | $1.00 +900% | $0.10 |
| Median input, per 1M | $2.00 +900% | $0.20 |
| Lowest output, per 1M | $5.00 +733.3% | $0.60 |
| Median output, per 1M | $10.00 +733.3% | $1.20 |
| Lowest unit price | — | — |
| Cost of 1M in + 200k out | $2.00 +809.1% | $0.22 |
| Context | 1M +4.8% | 1.1M |
| Max output | 128K | 128K |
| Providers | 33 | 28 +15.2% |
| Regions | APAC, AU, EU, GLOBAL, IN, JP, US | EU, GLOBAL, IN, US |
| Accepts | application/pdf, file, image, image/gif, image/jpeg, image/png, image/webp, pdf, text | application/pdf, file, image, image/gif, image/jpeg, image/png, image/webp, pdf, text |
| Returns | text | text |
| Type | Text models | Text models |
| Free option | no | no |
| Cheapest provider | openrouter | openrouter |
Your own workload: how many tokens the model gets (input) and gives back (output) - for example, in a month. The "Cost" row of the table above is recalculated for these numbers, for each model.
Computed from current offers each time the page opens; the best value of a row is bold, the percent is how far behind the best.
Providers selling both
302ai, abacus, aihubmix, amazon-bedrock, azure, azure-cognitive-services, cortecs, digitalocean, edenai, fastrouter, github-copilot, impossibl, jiekou, kenari, kilo, nearai, ofox, opencode, openrouter, orcarouter, perplexity-agent, requesty, standardcompute, trustedrouter, vercel, zenmux
Frequently asked questions
Which is cheaper, Claude Sonnet 5 or GPT-5.6 Luna?
GPT-5.6 Luna is 90% cheaper than Claude Sonnet 5 on input price. GPT-5.6 Luna is 88% cheaper than Claude Sonnet 5 on output price. GPT-5.6 Luna is 89% cheaper than Claude Sonnet 5 on a 1M-in / 200k-out workload.
Which has the larger context, Claude Sonnet 5 or GPT-5.6 Luna?
Claude Sonnet 5: 1M tokens; GPT-5.6 Luna: 1.1M tokens.
Where can I use both Claude Sonnet 5 and GPT-5.6 Luna?
302ai, abacus, aihubmix, amazon-bedrock, azure, azure-cognitive-services, cortecs, digitalocean, edenai, fastrouter, github-copilot, impossibl, jiekou, kenari, kilo, nearai, ofox, opencode, openrouter, orcarouter, perplexity-agent, requesty, standardcompute, trustedrouter, vercel, zenmux