Pairwise benchmark snapshot · Aug 17, 2026
In this head-to-head showdown, Claude Haiku 4.5 delivers higher overall intelligence and human-preferred responses (Elo 1,438 vs 1,425), while Qwen2.5 Coder 32B Instruct leads in SWE-bench software engineering benchmarks (65.2%), while Qwen2.5 Coder 32B Instruct is more budget-friendly at $1/1M per 1M output tokens. Review the complete breakdown below to determine which model best fits your performance and budget requirements.
7 benchmarks evaluated across capability, speed, and cost
Anthropic cheap/fast Claude SKU. Official Claude Platform list price $1/$5 per 1M tokens (Anthropic Haiku page).
Auto-discovered from OpenRouter (qwen/qwen-2.5-coder-32b-instruct). Preview until a second source matches.
Currently comparing Claude Haiku 4.5 vs Qwen2.5 Coder 32B Instruct. Alternative family tiers available.
Δ 13 Arena Elo pts
No latency data
$1 / 1M output
Relative percentile scores computed across all active models in the benchmark catalog.
Comparing top standard models across 6 key capability 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 Haiku 4.5 | Qwen2.5 Coder 32B Instruct | Advantage Delta |
|---|---|---|---|
| Preference Elo | +13 | ||
| Coding Elo | +49 | ||
| LiveBench | +2.4% | ||
| SWE-bench | +7% | ||
| GPQA Diamond | — | — | |
| Time to first token | — | — | |
| Output speed | 175 tok/s | — | — |
| Input price | $0.66/1M | +$0.34/1M | |
| Output price | +$4/1M | ||
| Context window | +167k |
Simulate monthly production API costs in USD (US Dollar).
Qwen2.5 Coder 32B Instruct is estimated to save $71.90/month ($863/year).
Higher SWE-bench (65.2%).
Lower output list price ($1/1M).
Need TTFT on both sides.
Larger window (200k).
Claude Haiku 4.5 is the side marked multimodal in the catalog.
| Target Workload | Recommended Pick | Evaluation Rationale |
|---|---|---|
| Repo / coding agents | Qwen2.5 Coder 32B Instruct | Higher SWE-bench (65.2%). |
| High-volume chat | Qwen2.5 Coder 32B Instruct | Lower output list price ($1/1M). |
| Voice / low-latency UI | insufficient data | Need TTFT on both sides. |
| Long-document RAG | Claude Haiku 4.5 | Larger window (200k). |
| Screenshots / vision | Claude Haiku 4.5 | Claude Haiku 4.5 is the side marked multimodal in the catalog. |
$1/$5 hosted Claude vs a cheap long-context open sibling. The alternative-to-Sonnet pair that is honest.
Huge window, cheap hosted lists. The Haiku alternative that is actually open-weight.
Official Anthropic list $1/$5. This is the SKU behind “cheapest Claude alternative” that is still Claude.
No comments posted on this matchup yet. Be the first to share an evaluation note!
Dynamic OpenGraph banner generated at /compare/claude-haiku-4-5-vs-qwen-2-5-coder-32b-instruct/opengraph-image
Verified Head-to-Head Telemetry
