Doubao Multi-Turn Bench
Models answer the turn of a real Doubao conversation where the user was dissatisfied, with the earlier turns as context. Rubrics grade each answer.
In-house Seed benchmark filtered from online Doubao user conversations.
- Made by
- ByteDance Seed
- Released
- June 2026
- Kind
- Multi-turn conversation
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- Kept private
- Strongest model tested
- GPT-5.5 #20 on the capability index
- Last in a launch post
- June 2026
- Last model added
- June 2026
- Scores refreshed
- November 12th, 2026
Seed2.1 Pro leads with 52.5%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
GPT-5.5 leads with 62.5% across all independent runs
Best score per model across the official leaderboard, Epoch AI’s own runs and any board Epoch AI copied. Each bar and row names its source.
- 1
Seed2.1 Pro 52.5%
Source: Labs’ model cards
- 1 GPT-5.5 62.5%
- 2 Seed2.1 Pro 52.5%
- 3
Gemini 3.1 Pro 52% - 4 Claude Opus 4.7 49.8%
Source: Official Doubao Multi-Turn Bench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Seed2.1 Pro | 52.5% | Jun 2026 | Lab model card |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5.5 | 62.5% | Apr 2026 | Official leaderboard | |
| 2 | Seed2.1 Pro | 52.5% | Jun 2026 | Official leaderboard | |
| 3 | Gemini 3.1 Pro | 52% | Feb 2026 | Official leaderboard | |
| 4 | Claude Opus 4.7 | 49.8% | Apr 2026 | Official leaderboard |
The benchmark's own leaderboard has 5 entries, 4 of them models we track. Switch the scores above to rank them together.
Number one over time Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Doubao Multi-Turn Bench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Sources: Official Doubao Multi-Turn Bench leaderboard and labs’ model cards
One lab put it in a launch post in the last six months
Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| Jun 23, 2026 | Seed2.1 Pro | 52.5 | Doubao Multi-Turn Bench |
One system card reports it
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Jun 23, 2026 | Seed2.1 Pro model card | 52.5 | Doubao Multi-Turn Bench (Table 2) | Capabilities |
One lab claim next to the leaderboard
Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. 1 of them match, which usually means one side copied the other. Every benchmark.
| Model | Lab says | Board says | Gap |
|---|---|---|---|
| Seed2.1 Pro | 52.5% Doubao Multi-Turn Bench | 52.5% Leaderboard | Same |
More in Instructions and chat
- IFEval In use Models answer about 500 prompts containing verifiable instructions such as word limits or required keywords. Code checks each instruction for strict and loose accuracy. Gemma 4 31B IT 98.9%
- IFBench In use 58 new out-of-domain verifiable output constraints. Scored by programmatic constraint checks (strict or loose).
Nemotron-Cascade 2 30B-A3B 82.9% - MultiChallenge In use Multi-turn conversations testing instruction retention, inference memory, versioned editing and self-coherence. Graded by LLM judge with rubrics. GPT-6.1 Sol 82%
- Multi-IF In use Models follow verifiable instructions over three-turn conversations in eight languages, extending IFEval. Scored by instruction accuracy at each turn. LFM3-1.7B 89.1%
- AdvancedIF In use Models respond to over 1,600 prompts with complex, multi-turn, and system-level instructions. A judge model checks each response against expert-written rubrics.
MAI-Thinking-1 85% - IFStruct In use Models generate JSON or YAML that must follow a requested schema and phrasing constraints, without constrained decoding. A response passes only if every structural constraint holds. Gemma 4 31B IT 95.9%