Claude Sonnet 4.6 vs Claude Haiku 4.5
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
| Specification |
by Anthropic
|
by Anthropic
|
|---|---|---|
| Overview | ||
| Status | Current mid tier | Current budget |
| Released | Jan 15, 2026 | Oct 15, 2025 |
| Pricing per million tokens | ||
| Input | $3.00/Mtok | $1.00/Mtok |
| Output | $15.00/Mtok | $5.00/Mtok |
| Blended avg | $6.00/Mtok | $2.00/Mtok |
| Cached input | $0.300/Mtok | $0.100/Mtok |
| Specifications | ||
| Context window | 1M tokens | 200K tokens |
| Parameters | Proprietary | Proprietary |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
textimage
|
textimage
|
| Benchmarks source: Vellum LLM Leaderboard | ||
| Avg benchmark score | 70.4 | 65.5 |
| Perf / dollar | 11.7 | 32.8 |
| GPQA Diamond | 89.9 | 73 |
| Terminal-Bench 2.1 | 59.1 | 27.5 |
| MCP Atlas | 69.5 | 40.2 |
| AIME 2025 | 83 | 96.3 |
| MMMLU | 89.3 | 83 |
| HumanEval | 79.6 | 73.3 |
| LiveCodeBench | 72.4 | — |
| BrowseComp | 76.2 | — |
| OSWorld-Verified | 78.5 | — |
| Humanity's Last Exam | 19.1 | — |
| ARC-AGI 2 | 58.3 | — |
| MATH 500 | — | 78 vendor |
| AutoBench | 5.3 | — |
| Independent composite scores each on its own scale — not part of the average above | ||
| AA Intelligence Index | 48.4 | 29.9 |
| AA Agentic Index | 42.1 | 16.5 |
| AA-Omniscience Index (−100–100) | 12.2 | -4.4 |
| LMSYS Chatbot Arena | 1472 ELO | 1412.5 ELO |
| Providers | ||
| Available from |
Anthropic — $3.00/$15.00/Mtok
|
Anthropic — $1.00/$5.00/Mtok
|
Cost at scale
1M tokens · 50/50 input/output| Volume | Claude Sonnet 4.6 | Claude Haiku 4.5 | Savings |
|---|---|---|---|
| 1M tokens | $6 | $2 | $4 (66.7%) |
| 10M tokens | $60 | $20 | $40 (66.7%) |
| 100M tokens | $600 | $200 | $400 (66.7%) |
| 1000M tokens | $6000 | $2000 | $4000 (66.7%) |
When to pick which
distilled from the pricing and spec data above- you need the longer context: 1M tokens vs 200K (5×)
- raw capability matters more than price: 70.4 vs 65.5 average benchmark score
- cost dominates: $2.00/Mtok blended vs $6.00 — 66.7% less on the same 50/50 token mix
The cheaper option here — Claude Haiku 4.5 costs $2.00/Mtok blended on Anthropic.
Summary
Claude Sonnet 4.6 by Anthropic costs $3.00/Mtok input and $15.00/Mtok output, with a 1M-token context window. It supports text, image input.
Claude Haiku 4.5 by Anthropic costs $1.00/Mtok input and $5.00/Mtok output, with a 200K-token context window. It supports text, image input.
On a blended cost basis, Claude Haiku 4.5 is 66.7% cheaper than Claude Sonnet 4.6.
On benchmarks, Claude Sonnet 4.6 scores higher (70.4 vs 65.5) on average. In terms of value, Claude Haiku 4.5 has better performance per dollar (32.8 vs 11.7).
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.