Claude Haiku 4.5 (Reasoning) vs Qwen 3.6 35B-A3B (Reasoning)
Results updated 5 November 2026 · In Claude Code and Qwen Code, so part of any gap can come from the harness · How we benchmark
Verdict
Claude Haiku 4.5 (Reasoning) edges Qwen 3.6 35B-A3B (Reasoning) by 1 point, and is faster; Qwen 3.6 35B-A3B (Reasoning) is cheaper.
- Cost per task
- $0.092 vs $0.00
- Time per task
- 1.6 min vs 2.9 min (estimated)
- Tokens per attempt
- 313k vs 650k
- Image input
- Yes vs Yes
Claude Haiku 4.5 (Reasoning), Claude Code scores higher on GeoBenchX and GS-QA2; Qwen 3.6 35B-A3B (Reasoning), Qwen Code scores higher on Earth-Bench V2. Qwen 3.6 35B-A3B (Reasoning), Qwen Code is cheaper ($0.00 against $0.092 per task), but Claude Haiku 4.5 (Reasoning), Claude Code is faster (1.6 min against 2.9 min).
Where they differ most
By kind of work, share of tasks passed, 0 to 100 · Higher is better
| Kind of work | Claude Haiku 4.5 (Reasoning) | Qwen 3.6 35B-A3B (Reasoning) |
|---|---|---|
| Climate and time series | 46 | 64 |
| GIS analysis | 67 | 56 |
| Making maps | 56 | 59 |
| Remote sensing and raster analysis | 64 | 65 |
Where the tasks are set: each kind of work links to its tasks and scoresCharts for each: Claude Haiku 4.5 (Reasoning), Claude Code · Qwen 3.6 35B-A3B (Reasoning), Qwen Code.
Side by side
| Metric | Claude Haiku 4.5 (Reasoning), Claude Code | Qwen 3.6 35B-A3B (Reasoning), Qwen Code |
|---|---|---|
| Geospatial Agent Index: Text only | 58 | 57 |
| Geospatial Agent Index: Text + vision | 48 | No data |
| Earth-Bench V2 score | 51 | 60 |
| GeoBenchX score | 53 | 50 |
| GS-QA2 score | 69 | 62 |
| Vision agents score | 19 | No data |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 78 | 65 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 74 | 47 |
| OGD4All score (standalone; in no index) | 95 | 93 |
| GeoNatureAgent score (standalone; in no index) | 93 | 79 |
| MultiGlobeQA score (standalone; in no index) | 70 | 72 |
| GeoAgentEval score (standalone; in no index) | 88 | 68 |
| ImpactMesh score (standalone; in no index) | No data | No data |
| Remote sensing and raster analysis score | 64 | 65 |
| Change detection score | 68 | 68 |
| GIS analysis score | 67 | 56 |
| Climate and time series score | 46 | 64 |
| Making maps score | 56 | 59 |
| Cost per task | $0.092 | $0.00 |
| Time per task (at 100 output tokens/s) | 1.6 min | 2.9 min (estimated) |
| Tokens per attempt | 313k | 650k |
| Turns per attempt | 10.6 | 16.9 |
| Input price per 1M tokens | $1.00 | $0.25 |
| Output price per 1M tokens | $5.00 | $1.49 |
| Context window | 200,000 | 262,144 |
| Image input | Yes | Yes |
| Developer | Anthropic | Alibaba (Qwen) |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, Claude Haiku 4.5 (Reasoning), Claude Code or Qwen 3.6 35B-A3B (Reasoning), Qwen Code?
Claude Haiku 4.5 (Reasoning), Claude Code, 58 to 57.
Which is cheaper per task, Claude Haiku 4.5 (Reasoning), Claude Code or Qwen 3.6 35B-A3B (Reasoning), Qwen Code?
Qwen 3.6 35B-A3B (Reasoning), Qwen Code, $0.00 to $0.092.
Which writes fewer output tokens per task, Claude Haiku 4.5 (Reasoning), Claude Code or Qwen 3.6 35B-A3B (Reasoning), Qwen Code?
Claude Haiku 4.5 (Reasoning), Claude Code, 7k to 18k.
Which has the larger context window, Claude Haiku 4.5 (Reasoning), Claude Code or Qwen 3.6 35B-A3B (Reasoning), Qwen Code?
Qwen 3.6 35B-A3B (Reasoning), Qwen Code, 262,144 tokens to 200,000 tokens.