Claude Haiku 4.5
Last updated: August 10, 2026
Specs
| Vendor | Anthropic |
|---|
| Released | October 1, 2025 |
|---|
| Context window | 200K tokens |
|---|
| Input | $1 / M tokens |
|---|
| Output | $5 / M tokens |
|---|
Benchmarks
| mmlu | 81.5 |
|---|
| gpqa | 48 |
|---|
| humanEval | 78.2 |
|---|
| arenaElo | 1274 |
|---|
| sweBench | 40.6 |
|---|
| sweBenchPro | 32 |
|---|
| liveCodeBench | 55 |
|---|
| aiderPolyglot | 52 |
|---|
| terminalBench | 28 |
|---|
| mmluPro | 68 |
|---|
| hle | 9 |
|---|
| aime | 60 |
|---|
| math500 | 84 |
|---|
| mmmu | 64 |
|---|
Strengths
- Extremely low latency
- Cheapest Anthropic model
- Great for high-volume classification and routing
Weaknesses
- Not suited for deep reasoning tasks
- Lower coding accuracy on complex problems
Best for
- High-volume support chatbots
- Content moderation and classification
- Realtime voice agents
Head-to-head comparisons
Compare with other AI models