Pairwise benchmark snapshot · Aug 1, 2026
In this head-to-head showdown, Qwen 2.5 Coder 32B delivers higher overall intelligence and human-preferred responses (Elo 1,425 vs 1,224), while Qwen 2.5 Coder 32B leads in SWE-bench software engineering benchmarks (65.2%), while Qwen 2.5 Coder 32B is more budget-friendly at $0.72/1M per 1M output tokens, while Claude 3.5 Haiku provides faster response streaming (140 tok/s). Review the complete breakdown below to determine which model best fits your performance and budget requirements.
9 direct benchmark disciplines evaluated across capability, speed, and cost
Previous cheap Claude. Haiku 4.5 is the current $1/$5 SKU.
Alibaba dedicated open-weight code generation model with near-frontier SWE-bench Verified coding capability.
Δ 201 Arena Elo pts
140 tok/s
$0.72 / 1M output
Relative percentile scores computed across all active models in the benchmark catalog.
Comparing top Western standard models with China's leading frontier rival across 6 skill dimensions. Tap any spoke or dot to inspect.
Category wins across reasoning intelligence, generation speed, and token cost.
Preference Elo, Coding proficiency, SWE-bench & LiveBench accuracy
Generation throughput and time to first token responsiveness
Cost per million tokens and max context window length
| Benchmark | Claude 3.5 Haiku | Qwen 2.5 Coder 32B | Advantage Delta |
|---|---|---|---|
| Preference Elo | +201 | ||
| Coding Elo | — | — | |
| LiveBench | +20% | ||
| SWE-bench | +29% | ||
| GPQA Diamond | +19.9% | ||
| Time to first token | +70 ms | ||
| Output speed | 140 tok/s | 110 tok/s | +30 tok/s |
| Input price | $0.8/1M | $0.18/1M | +$0.62/1M |
| Output price | $0.72/1M | +$3.28/1M | |
| Context window | +69k |
Simulate monthly production API costs in USD (US Dollar).
Qwen 2.5 Coder 32B is estimated to save $70.90/month ($851/year).
Higher SWE-bench (65.2%).
Lower output list price ($0.72/1M).
Lower TTFT (110 ms).
Larger window (200k).
Claude 3.5 Haiku is the side marked multimodal in the catalog.
| Target Workload | Recommended Pick | Evaluation Rationale |
|---|---|---|
| Repo / coding agents | Qwen 2.5 Coder 32B | Higher SWE-bench (65.2%). |
| High-volume chat | Qwen 2.5 Coder 32B | Lower output list price ($0.72/1M). |
| Voice / low-latency UI | Claude 3.5 Haiku | Lower TTFT (110 ms). |
| Long-document RAG | Claude 3.5 Haiku | Larger window (200k). |
| Screenshots / vision | Claude 3.5 Haiku | Claude 3.5 Haiku is the side marked multimodal in the catalog. |
No comments posted on this matchup yet. Be the first to share an evaluation note!
Dynamic OpenGraph banner generated at /compare/claude-3-5-haiku-vs-qwen-2-5-coder-32b/opengraph-image
Verified Head-to-Head Telemetry
