Skip to content

Claude Haiku 5.5

AnthropicProprietaryReasoningProvisionalReleased Oct 7, 2026Input: text, image, file
Compare with MiniMax M3.1 FlashSource scores ↗OpenRouter ↗
Overall score
67.2
Rank 26 of 222 · 52.9–81.6
Context window
1M
1,000,000 tokens
Price per 1M tokens
$0.10 in · $0.50 out
✓ Verified on Anthropic's pricing page
Blended price
$0.20
3:1 input:output tokens
Output speed
—
Not measured yet

Category scores

Scores per capability area with the model's rank among ranked models in that category.

  1. 1Math79.1
  2. 2Knowledge#2967.3
  3. 3Coding#4164.8
  4. 4Reasoning#2360.1
Higher is better. #N is the model's rank in that category, where it is eligible to be ranked.

Benchmark results (18)

As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.

Agentic & tool use

MazeBench1#13 of 42
HLE w/ tools57.4#8 of 27
Terminal-Bench 4.039.20#11 of 18
Hermes Index37.0#10 of 16
Hermes Bench71.0#9 of 16

Reasoning

ARC-AGI-261.3#25 of 41
ARC-AGI-189.00#23 of 36
ARC-AGI-30.5#13 of 22
ARC-AGI-3 Provider Adapter2.82#7 of 8

Coding

Bug Hunt Bench21.5#17 of 23
FrontierSWE v243.8#8 of 20
FrontierCode 1.1 Main46.4#7 of 16

Multimodal

Design Arena Website1263#34 of 91
ParseBench57.40#30 of 37
Chartography (no tools)46.4#4 of 5

Math

FrontierMath v2 (Tiers 1-3)75.088#19 of 64
FrontierMath v2 (Tier 4)46.341#19 of 51

Knowledge

HLE w/o tools45.9#10 of 43

Pricing by source

USD per 1M tokens.

SourceInputOutputCached in
✓ Official (Anthropic)
checked Oct 17, 2026 · Claude Haiku 5.5
$0.10$0.50$0.010
BenchLM list price$0.10$0.50$0.010
OpenRouter anthropic/claude-haiku-5.5$0.10$0.50$0.010

Prompts up to 100K tokens; $0.50 / $2.50 above.

Anthropic lists $0.10 input / $0.50 output per million tokens for prompts up to 100,000 tokens; prompts over 100,000 tokens cost $0.50 / $2.50. Cache reads cost $0.01 / $0.05 per million tokens at those respective tiers; cache writes cost $0.125 / $0.625 (5-minute) or $0.20 / $1.00 (1-hour). Batch API input and output receive a 50% discount. Pricing fields store the lower prompt tier.

Arena ratings

Human-preference ratings from LMArena. Best-rated variant shown.

WebDev (coding)(High)1586#34 · 3.5K votes

Speed

Measured by us through OpenRouter (median of trials).

Not measured yet.

Related models

Same family

Similar score