Compare · ModelsLive · 2 picked · head to head
Claude Haiku 5.5 vs Qwen3.6 27B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Claude Haiku 5.5 wins on 6/6 benchmarks
Claude Haiku 5.5 wins 6 of 6 shared benchmarks. Leads in speed · knowledge · math.
Category leads
speed·Claude Haiku 5.5knowledge·Claude Haiku 5.5math·Claude Haiku 5.5general·Claude Haiku 5.5
Hype vs Reality
Attention vs performance
Claude Haiku 5.5
#109 by perf·#15 by attention
Qwen3.6 27B
#68 by perf·#2 by attention
Best value
Claude Haiku 5.5
5.4x better value than Qwen3.6 27B
Claude Haiku 5.5
175.3 pts/$
$0.30/M
Qwen3.6 27B
32.3 pts/$
$1.76/M
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
6 benchmarks · 2 models
Claude Haiku 5.5Qwen3.6 27B
Artificial Analysis · Quality Index
Claude Haiku 5.5 leads by +6.4
Claude Haiku 5.5
43.4
Qwen3.6 27B
37.0
Chess Puzzles
Claude Haiku 5.5 leads by +34.7
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Claude Haiku 5.5
52.6
Qwen3.6 27B
17.9
FrontierMath-Tiers-1-3-v2-Private
Claude Haiku 5.5 leads by +40.0
Claude Haiku 5.5
75.1
Qwen3.6 27B
35.1
GPQA diamond
Claude Haiku 5.5 leads by +5.1
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude Haiku 5.5
86.2
Qwen3.6 27B
81.1
Mystery Game Puzzles
Claude Haiku 5.5 leads by +22.9
Claude Haiku 5.5
22.9
Qwen3.6 27B
0.0
OTIS Mock AIME 2024-2025
Claude Haiku 5.5 leads by +7.8
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Haiku 5.5
98.9
Qwen3.6 27B
91.1
Full benchmark table
| Benchmark | Claude Haiku 5.5 | Qwen3.6 27B |
|---|---|---|
Artificial Analysis · Quality Index | 43.4 | 37.0 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 52.6 | 17.9 |
FrontierMath-Tiers-1-3-v2-Private | 75.1 | 35.1 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 86.2 | 81.1 |
Mystery Game Puzzles | 22.9 | 0.0 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 98.9 | 91.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.10 | $0.50 | 1.0M tokens (~500 books) | $2.00 | |
| $0.32 | $3.20 | 262K tokens (~131 books) | $10.40 |