UserBench (developer simulation)
Models predict a developer’s next message in a real coding session. The evaluator compares predicted and observed action labels using Jaccard similarity.
On November 19, 2026, GPT-5.6 Sol at high effort scored 50.0% with 400 prior turns and 45.9% without history across 580 scored tasks.
- Made by
- UserBench team
- Released
- July 2026
- Kind
- Multi-turn conversation
- Level
- University
- Scored
- Checked automatically
- Answers
- Prediction
- Tasks
- All public
- Measures
- Capability
- Language
- English
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- Occupations
- Computer and mathematical
- Strongest model tested
- GPT-5.6 Sol #14 on the capability index
- Last model added
- July 2026
- Scores refreshed
- November 20th, 2026
GPT-5.6 Sol leads with 51.3%
Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best.
- 1 GPT-5.6 Sol train400 / max 51.3%
- 2
DeepSeek V4 Flash 0731 train400 / high 46.9%
Source: Official UserBench (developer simulation) leaderboard
| # | Model | Score | Setting | Date | Link | Source |
|---|---|---|---|---|---|---|
| 1 | GPT-5.6 Sol | 51.3% | train400 / max | Jul 2026 | Official leaderboard | |
| 2 | | 46.9% | train400 / high | Jul 2026 | hub.harborframework.com | Official leaderboard |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official UserBench (developer simulation) leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official UserBench (developer simulation) leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
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). Flint 89.2%
- 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%
- MM-IFEval In use Multimodal models follow image-grounded instructions with exact output constraints. Checks mix rule-based verification and a judge model. GPT-4o 64.6%
Data audit
All 9 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (2 entries).
- Note Models scored Only 2 models we track.
- Pass Newest score The newest is from July 2026, for DeepSeek V4 Flash 0731.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.6 Sol (#14).
- Pass Headroom The best is 51.3%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on userbench.vercel.app: 11 rows for models we track at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.