DeepSeek complex instruction following and multi-turn writing evaluation
DeepSeek-V4-Pro 45.9% Models answer 196 hard Chinese writing prompts with complex constraints or several turns. Pairwise judging against a baseline model gives a win rate.
In-house DeepSeek evaluation. The prompts are not public.
- Made by
- DeepSeek
- Released
- April 2026
- Kind
- Multi-turn conversation
- Level
- Everyday
- Scored
- Ranked by matches
- Answers
- Open-ended
- Tasks
- Kept private
- Strongest model tested
- DeepSeek-V4-Pro #69 on the capability index
- Last in a system card
- April 2026
- Last model added
- April 2026
Only DeepSeek-V4-Pro has a rating here, from its lab’s own post
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. A rating ranks a model only against others rated in the same arena at the same time, so these don’t show who leads.
- 1 DeepSeek-V4-Pro 45.9%
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | DeepSeek-V4-Pro | 45.9% | Apr 2026 | Lab model card |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Labs’ model cards
Labs report it only in system cards
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark. Some system cards do, below.
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 |
|---|---|---|---|---|
| Apr 26, 2026 | DeepSeek-V4-Pro technical report | 45.9 | Complex Instruction Following and Multi-Turn Writing, total DS win rate vs Claude-Opus-4.5, Opus 52.0, tie 2.0 (Table 14) | Capabilities |
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%