Model
Qwen3.8 Max Preview
Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows
Compare providers (sorted by listed input price)
3 providers list Qwen3.8 Max Preview. The cheapest full-context option is Charm Hyper at $2/$6 per M.
Best price history
Daily lowest published price across live providers. A lower line is better for the buyer.
- Best input /M
- $2▲ 10.2%
- Best output /M
- $6▲ 10.2%
- Provider coverage
- 3
- 5 at the start · 24 snapshots
Price badge
https://www.vaanalytics.in/badge/alibaba/qwen3.8-max-preview.svgLive SVG, regenerated on every hourly sync — paste it into a README to always show the current cheapest listed price. Updates when prices change; no build step needed on your side.
Recent changes
Changelog →- deprecatedvia alibaba-token-plan-cn · Sep 11, 2026
- status: beta → deprecated
- deprecatedvia alibaba-token-plan · Sep 11, 2026
- status: beta → deprecated
- repricedvia llmgateway · Sep 10, 2026
- input: $1.81 → $2
- output: $5.45 → $6
- cache_read: $0.21 → $0.25
Benchmarks
- Terminal-Bench ↗(accuracy)report ↗Unclassified source86.6
- SWE-Bench Pro ↗(resolve rate)report ↗Unclassified source67.7
- DeepSWE(resolve rate)report ↗Unclassified source56.6
- NL2Repo(resolve rate)report ↗Unclassified source55.9
- FrontierSWE ↗(dominance score)report ↗Unclassified source73.5
- MLS-Bench-Lite(score)report ↗Unclassified source41
- AutomationBench(pass@1)report ↗Unclassified source27.3
- Toolathlon Verified(pass@1)report ↗Unclassified source72.5
- WideSearch(F1)report ↗Unclassified source81.9
- Humanity's Last Exam ↗(accuracy)report ↗Unclassified source56.2
- GPQA Diamond ↗(accuracy)report ↗Unclassified source92.6
- Humanity's Last Exam ↗(accuracy)report ↗Unclassified source43.6
- IFBench(score)report ↗Unclassified source82.8
- OSWorld-Verified ↗(success rate)report ↗Unclassified source86.1
- MMMU Pro ↗(accuracy)report ↗Unclassified source82.3
Scores a lab published about its own model are claims, not measurements. Rankings elsewhere on this site use independently measured scores only.
Prices are per million tokens (USD). “—” means the provider does not publicly list a price for this model. Data from models.dev; verify with the provider before purchasing.