# Qwen3.5 397B A17BThinking

> Source: Benchize (https://benchize.com) — https://benchize.com/models/qwen3-5-397b-a17b-thinking. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/models/qwen3-5-397b-a17b-thinking.

Alibaba

Canonical URL: https://benchize.com/models/qwen3-5-397b-a17b-thinking
Last updated: Thu, 22 Oct 2026 06:30:09 GMT

Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations.

Alibaba

## Benchmarks

| Benchmark | Score | Rank |  |
| --- | --- | --- | --- |
| [τ-bench banking knowledge](https://benchize.com/benchmarks/sierra%3Abanking_knowledge.md) | 9.79% | 19 / 19 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `pass_1` \| 9.79 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_2` \| 6.36 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_3` \| 5.41 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_4` \| 5.15 \| percent \| consistent_success_across_k_trials \| Higher is better \| ### Configuration benchmarkVersion 1.0.1 harness tau-bench methodology {"evaluation_date":"2026-02-27","notes":"Evaluated using Qwen3.5-397B-A17B with thinking enabled. User simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Seed: 300. Temperature: 1.0, top_p: 0.95. Self-hosted via vLLM. banking_knowledge re-graded under tau2-bench v1.0.1 grading fixes on 2026-07-15 (extra-read logging #329, numeric-argument normalization #397, T077-T086 gold trajectories #402); scores prior to v1.0.1 are not comparable.","tau2_bench_version":"1.0.1","user_simulator":"gpt-5.2","verification":{"details":"Standard tau-bench scaffold. Clean 4-trial runs with no infrastructure errors across all domains.","modified_prompts":false,"omitted_questions":false}} modality text Reasoning effort enabled retrieval text-emb-3-large sourceSubmission qwen3.5-397b-a17b-think_sierra_2026-03-02 ### Source [Original source](https://sierra-tau-bench-public.s3.us-west-2.amazonaws.com/submissions/qwen3.5-397b-a17b-think_sierra_2026-03-02/submission.json) Source sierra Original model label Qwen3.5-397B-A17B Evaluated 2026-03-02 Observed 2026-10-21T16:54:48.177Z Evidence SHA-256 `2cfb2c7d2a46e9f22154fa153dcdf155d50dbd5bf86233ecf564eccca1e5913e` Source locator `/results/banking_knowledge` |  |  |  |
| [τ-bench retail](https://benchize.com/benchmarks/sierra%3Aretail.md) | 84.43% | 1 / 8 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `pass_1` \| 84.43 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_2` \| 74.42 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_3` \| 66.45 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_4` \| 59.65 \| percent \| consistent_success_across_k_trials \| Higher is better \| ### Configuration benchmarkVersion 1.0.1 harness tau-bench methodology {"evaluation_date":"2026-02-27","notes":"Evaluated using Qwen3.5-397B-A17B with thinking enabled. User simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Seed: 300. Temperature: 1.0, top_p: 0.95. Self-hosted via vLLM. banking_knowledge re-graded under tau2-bench v1.0.1 grading fixes on 2026-07-15 (extra-read logging #329, numeric-argument normalization #397, T077-T086 gold trajectories #402); scores prior to v1.0.1 are not comparable.","tau2_bench_version":"1.0.1","user_simulator":"gpt-5.2","verification":{"details":"Standard tau-bench scaffold. Clean 4-trial runs with no infrastructure errors across all domains.","modified_prompts":false,"omitted_questions":false}} modality text Reasoning effort enabled retrieval Not reported sourceSubmission qwen3.5-397b-a17b-think_sierra_2026-03-02 ### Source [Original source](https://sierra-tau-bench-public.s3.us-west-2.amazonaws.com/submissions/qwen3.5-397b-a17b-think_sierra_2026-03-02/submission.json) Source sierra Original model label Qwen3.5-397B-A17B Evaluated 2026-03-02 Observed 2026-10-21T16:54:48.177Z Evidence SHA-256 `2cfb2c7d2a46e9f22154fa153dcdf155d50dbd5bf86233ecf564eccca1e5913e` Source locator `/results/retail` |  |  |  |
| [τ-bench airline](https://benchize.com/benchmarks/sierra%3Aairline.md) | 81.5% | 4 / 8 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `pass_1` \| 81.5 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_2` \| 75.67 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_3` \| 71.5 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_4` \| 68 \| percent \| consistent_success_across_k_trials \| Higher is better \| ### Configuration benchmarkVersion 1.0.1 harness tau-bench methodology {"evaluation_date":"2026-02-27","notes":"Evaluated using Qwen3.5-397B-A17B with thinking enabled. User simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Seed: 300. Temperature: 1.0, top_p: 0.95. Self-hosted via vLLM. banking_knowledge re-graded under tau2-bench v1.0.1 grading fixes on 2026-07-15 (extra-read logging #329, numeric-argument normalization #397, T077-T086 gold trajectories #402); scores prior to v1.0.1 are not comparable.","tau2_bench_version":"1.0.1","user_simulator":"gpt-5.2","verification":{"details":"Standard tau-bench scaffold. Clean 4-trial runs with no infrastructure errors across all domains.","modified_prompts":false,"omitted_questions":false}} modality text Reasoning effort enabled retrieval Not reported sourceSubmission qwen3.5-397b-a17b-think_sierra_2026-03-02 ### Source [Original source](https://sierra-tau-bench-public.s3.us-west-2.amazonaws.com/submissions/qwen3.5-397b-a17b-think_sierra_2026-03-02/submission.json) Source sierra Original model label Qwen3.5-397B-A17B Evaluated 2026-03-02 Observed 2026-10-21T16:54:48.177Z Evidence SHA-256 `2cfb2c7d2a46e9f22154fa153dcdf155d50dbd5bf86233ecf564eccca1e5913e` Source locator `/results/airline` |  |  |  |
| [τ-bench telecom](https://benchize.com/benchmarks/sierra%3Atelecom.md) | 97.81% | 1 / 8 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `pass_1` \| 97.81 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_2` \| 95.76 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_3` \| 93.86 \| percent \| consistent_success_across_k_trials \| Higher is better \| \| `pass_4` \| 92.11 \| percent \| consistent_success_across_k_trials \| Higher is better \| ### Configuration benchmarkVersion 1.0.1 harness tau-bench methodology {"evaluation_date":"2026-02-27","notes":"Evaluated using Qwen3.5-397B-A17B with thinking enabled. User simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Seed: 300. Temperature: 1.0, top_p: 0.95. Self-hosted via vLLM. banking_knowledge re-graded under tau2-bench v1.0.1 grading fixes on 2026-07-15 (extra-read logging #329, numeric-argument normalization #397, T077-T086 gold trajectories #402); scores prior to v1.0.1 are not comparable.","tau2_bench_version":"1.0.1","user_simulator":"gpt-5.2","verification":{"details":"Standard tau-bench scaffold. Clean 4-trial runs with no infrastructure errors across all domains.","modified_prompts":false,"omitted_questions":false}} modality text Reasoning effort enabled retrieval Not reported sourceSubmission qwen3.5-397b-a17b-think_sierra_2026-03-02 ### Source [Original source](https://sierra-tau-bench-public.s3.us-west-2.amazonaws.com/submissions/qwen3.5-397b-a17b-think_sierra_2026-03-02/submission.json) Source sierra Original model label Qwen3.5-397B-A17B Evaluated 2026-03-02 Observed 2026-10-21T16:54:48.177Z Evidence SHA-256 `2cfb2c7d2a46e9f22154fa153dcdf155d50dbd5bf86233ecf564eccca1e5913e` Source locator `/results/telecom` |  |  |  |

---
Cite as: Qwen3.5 397B A17BThinking — Benchize, https://benchize.com/models/qwen3-5-397b-a17b-thinking (retrieved 2026-10-22)