Anthropic · released 2026-04-08 · text model
Claude 4.6 Sonnet
AA INDEX89editorial 2026-05editorial estimate
Goes to OpenRouter · referral link · See disclosure →
Input price$3.00/1M tokens
Output price$15.00/1M tokens
Context200Ktokens
P50 latency1.1sfirst tokvendor-reported
Throughput85tok/svendor-reported
The strongest non-OpenAI option for nuanced writing, code review, and analysis that rewards careful language. Anthropic's refusal patterns are more reliable than mainstream alternatives for regulated industries (legal, healthcare, finance). Pricier than DeepSeek but the response quality on edge-case prompts is hard to replicate at a lower tier.
Full pricing details
all prices in USD per 1M tokens · All hidden costs surfaced| Pricing tier | Input | Output | Notes |
|---|---|---|---|
Standard Pay-as-you-go API rate card | $3.00 | $15.00 | Default if no headers set |
With prompt caching Cached read · cache write | $0.30−90% | $15.00 | Cache writes priced at standard input ($3.00); cached reads 90% off. Requires cache_control header. |
Batch API (50% off) 24h SLA, async | $1.50−50% | $7.50−50% | Send up to 50K requests in one batch; results within 24h |
Image input (vision) Per image, low/high detail differ | $3.00 + image fee | $15.00 | $0.50 per 1024×1024 image at high detail; low detail ~$0.10 |
Reasoning tokens Hidden chain-of-thought tokens | $3.00 | $15.00 × ~3-8 | When using reasoning.effort: high, output tokens include hidden reasoning. Real cost can be 3-8× headline output. |
Structured output JSON mode + schema | $3.00 | $15.00 | No extra cost; first call may be slower (~2x latency) due to schema compilation |
Capabilities
Vision (image understanding)Function callingStructured output (JSON schema)Reasoning modeMultilingual (32+ langs)Batch APIAudio I/OFine-tuning