EnterpriseOps-Gym
The official leaderboard hasn’t scored a model released after Gemini 3.1 Pro (February 2026). Labs report their own numbers for 4 newer models, such as Inkling 2 Small at 46.8%.
Agents do 1,150 expert-written tasks in eight enterprise domains, such as IT, HR and customer service, using 512 tools over live MCP servers. SQL checks on the final database state score each task.
Some tasks are impossible and test refusal. Artificial Analysis runs its own version named EnterpriseOps-Gym-AA.
- Made by
- ServiceNow
- Released
- March 2026
- Kind
- Apps and MCP servers
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Agent task
- Tasks
- All public
- Measures
- Capability
- Subject
- Economics and business
- Language
- English
- Interaction
- Agentic
- Environment
- APIs and tools
- Items from
- Written by experts
- Refresh
- New versions replace it
- Run by
- Anyone
- In views
- Agents, Law and finance
- Strongest model tested
- Claude Opus 4.6 #39 on the capability index
- Last in a launch post
- November 2026
- Last model added
- November 2026
- Scores refreshed
- November 18th, 2026
In the authors’ words
Large language models are shifting from passive information providers to active agents intended for complex workflows. However, their deployment as reliable AI workers in enterprise is stalled by benchmarks that fail to capture the intricacies of professional environments, specifically, the need for long-horizon planning amidst persistent state changes and strict access protocols. In this work, we introduce EnterpriseOps-Gym, a benchmark designed to evaluate agentic planning in realistic enterprise settings. Specifically, EnterpriseOps-Gym features a containerized sandbox with 164 database tables and 512 functional tools to mimic real-world search friction. Within this environment, agents are evaluated on 1,150 expert-curated tasks across eight mission-critical verticals (including Customer Service, HR, and IT). Our evaluation of 14 frontier models reveals critical limitations in state-of-the-art models: the top-performing Claude Opus 4.5 achieves only 37.4% success. Further analysis shows that providing oracle human plans improves performance by 14-35 percentage points, pinpointing strategic reasoning as the primary bottleneck. Additionally, agents frequently fail to refuse infeasible tasks (best model achieves 53.9%), leading to unintended and potentially harmful side effects. Our findings underscore that current agents are not yet ready for autonomous enterprise deployment. More broadly, EnterpriseOps-Gym provides a concrete testbed to advance the robustness of agentic planning in professional workflows.
Claude Opus 4.6 leads with 45.9% among models released by February 2026 Claude Opus 4.6 leads with 45.9% among models released by February 2026
Best score per entry from the benchmark's own leaderboard, read on November 18th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.
Inkling 2 Small leads with 46.8%, claims included
Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model. Striped bars are claims, and dotted bars are reruns by Epoch AI or another board.
The newest model with a score here came out in February 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 45.9 Claude Opus 4.6
- 42.2 Claude Sonnet 4.6
- 39.4 Claude Opus 4.5
- 38.7
Gemini 3.1 Pro
- 34.1 Claude Sonnet 4.5
- 34.0
Gemini 3 Flash
- 30.9 GPT-5
- 29.7
Gemini 3 Pro
- 24.2
DeepSeek-V3.2
- 22.9 GPT-5 mini
- 22.4 Kimi K2 Thinking
- 20.0 Gemini-2.5-Pro (snapshot unspecified)
- 19.1 Qwen3-30B
- 18.1 Qwen3-235B
- 15.3
Qwen3-4B
- 1 Claude Opus 4.6 Oracle 45.9%
- 2 Claude Sonnet 4.6 Oracle 42.2%
- 3 Claude Opus 4.5 Oracle 39.4%
- 4
Gemini 3.1 Pro Oracle 38.7%
- 5 Claude Sonnet 4.5 Oracle 34.1%
- 6
Gemini 3 Flash Oracle 34%
- 7 GPT-5 Oracle 30.9%
- 8
Gemini 3 Pro Oracle 29.7%
- 9
DeepSeek-V3.2 Oracle 24.2%
- 10 GPT-5 mini Oracle 22.9%
- 11 Kimi K2 Thinking Oracle 22.4%
- 12 Gemini-2.5-Pro (snapshot unspecified) Oracle 20%
- 13 Qwen3-30B Oracle; Think 19.1%
- 14 Qwen3-235B Oracle; Inst. 18.1%
- 15
Qwen3-4B Oracle; Think 15.3%
Source: Official EnterpriseOps-Gym leaderboard
- 1 Claude Opus 4.6 Oracle 45.9%
- 2 Claude Sonnet 4.6 Oracle 42.2%
- 3 Claude Opus 4.5 Oracle 39.4%
- 4
Gemini 3.1 Pro Oracle 38.7%
- 5 Claude Sonnet 4.5 Oracle 34.1%
- 6
Gemini 3 Flash Oracle 34%
- 7 GPT-5 Oracle 30.9%
- 8
Gemini 3 Pro Oracle 29.7%
- 9
DeepSeek-V3.2 Oracle 24.2%
- 10 GPT-5 mini Oracle 22.9%
- 11 Kimi K2 Thinking Oracle 22.4%
- 12
Qwen3-4B Oracle; Think 15.3%
Source: Official EnterpriseOps-Gym leaderboard
- 46.8
Inkling 2 Small
- 45.9 Claude Opus 4.6
- 42.2 Claude Sonnet 4.6
- 42.0
Inkling 2 Mini
- 39.4 Claude Opus 4.5
- 38.7
Gemini 3.1 Pro
- 34.1 Claude Sonnet 4.5
- 34.0
Gemini 3 Flash
- 33.6
Spark-X2.6-4B
- 30.9 GPT-5
- 29.7
Gemini 3 Pro
- 28.2 LFM3-1.7B
- 24.2
DeepSeek-V3.2
- 22.9 GPT-5 mini
- 22.4 Kimi K2 Thinking
- 15.3
Qwen3-4B
- 1
Inkling 2 Small 46.8%
- 2 Claude Opus 4.6 Oracle 45.9%
- 3 Claude Sonnet 4.6 Oracle 42.2%
- 4
Inkling 2 Mini 42%
- 5 Claude Opus 4.5 Oracle 39.4%
- 6
Gemini 3.1 Pro Oracle 38.7%
- 7 Claude Sonnet 4.5 Oracle 34.1%
- 8
Gemini 3 Flash Oracle 34%
- 9
Spark-X2.6-4B 33.6%
- 10 GPT-5 Oracle 30.9%
- 11
Gemini 3 Pro Oracle 29.7%
- 12 LFM3-1.7B 28.2%
- 13
DeepSeek-V3.2 Oracle 24.2%
- 14 GPT-5 mini Oracle 22.9%
- 15 Kimi K2 Thinking Oracle 22.4%
- 16
Qwen3-4B Oracle; Think 15.3%
Sources: Labs’ model cards and official EnterpriseOps-Gym leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.6 | 45.9% | Oracle | Feb 2026 | Official leaderboard |
| 2 | Claude Sonnet 4.6 | 42.2% | Oracle | Feb 2026 | Official leaderboard |
| 3 | Claude Opus 4.5 | 39.4% | Oracle | Nov 2025 | Official leaderboard |
| 4 | | 38.7% | Oracle | Feb 2026 | Official leaderboard |
| 5 | Claude Sonnet 4.5 | 34.1% | Oracle | Sep 2025 | Official leaderboard |
| 6 | | 34% | Oracle | Dec 2025 | Official leaderboard |
| 7 | GPT-5 | 30.9% | Oracle | Aug 2025 | Official leaderboard |
| 8 | | 29.7% | Oracle | Nov 2025 | Official leaderboard |
| 9 | | 24.2% | Oracle | Dec 2025 | Official leaderboard |
| 10 | GPT-5 mini | 22.9% | Oracle | Aug 2025 | Official leaderboard |
| 11 | Kimi K2 Thinking | 22.4% | Oracle | Nov 2025 | Official leaderboard |
| 12 | Gemini-2.5-Pro (snapshot unspecified) | 20% | Oracle | Official leaderboard | |
| 13 | Qwen3-30B | 19.1% | Oracle; Think | Official leaderboard | |
| 14 | Qwen3-235B | 18.1% | Oracle; Inst. | Official leaderboard | |
| 15 | | 15.3% | Oracle; Think | Apr 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.6 | 45.9% | Oracle | Feb 2026 | Official leaderboard |
| 2 | Claude Sonnet 4.6 | 42.2% | Oracle | Feb 2026 | Official leaderboard |
| 3 | Claude Opus 4.5 | 39.4% | Oracle | Nov 2025 | Official leaderboard |
| 4 | | 38.7% | Oracle | Feb 2026 | Official leaderboard |
| 5 | Claude Sonnet 4.5 | 34.1% | Oracle | Sep 2025 | Official leaderboard |
| 6 | | 34% | Oracle | Dec 2025 | Official leaderboard |
| 7 | GPT-5 | 30.9% | Oracle | Aug 2025 | Official leaderboard |
| 8 | | 29.7% | Oracle | Nov 2025 | Official leaderboard |
| 9 | | 24.2% | Oracle | Dec 2025 | Official leaderboard |
| 10 | GPT-5 mini | 22.9% | Oracle | Aug 2025 | Official leaderboard |
| 11 | Kimi K2 Thinking | 22.4% | Oracle | Nov 2025 | Official leaderboard |
| 12 | | 15.3% | Oracle; Think | Apr 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | 46.8% | Nov 2026 | Lab model card | ||
| 2 | Claude Opus 4.6 | 45.9% | Oracle | Feb 2026 | Official leaderboard |
| 3 | Claude Sonnet 4.6 | 42.2% | Oracle | Feb 2026 | Official leaderboard |
| 4 | 42% | Nov 2026 | Lab model card | ||
| 5 | Claude Opus 4.5 | 39.4% | Oracle | Nov 2025 | Official leaderboard |
| 6 | 38.7% | Oracle | Feb 2026 | Official leaderboard | |
| 7 | Claude Sonnet 4.5 | 34.1% | Oracle | Sep 2025 | Official leaderboard |
| 8 | 34% | Oracle | Dec 2025 | Official leaderboard | |
| 9 | 33.6% | Oct 2026 | Lab model card | ||
| 10 | GPT-5 | 30.9% | Oracle | Aug 2025 | Official leaderboard |
| 11 | 29.7% | Oracle | Nov 2025 | Official leaderboard | |
| 12 | LFM3-1.7B | 28.2% | Oct 2026 | Lab model card | |
| 13 | 24.2% | Oracle | Dec 2025 | Official leaderboard | |
| 14 | GPT-5 mini | 22.9% | Oracle | Aug 2025 | Official leaderboard |
| 15 | Kimi K2 Thinking | 22.4% | Oracle | Nov 2025 | Official leaderboard |
| 16 | 15.3% | Oracle; Think | Apr 2025 | Official leaderboard |
Labs report scores for 4 models in their cards. Switch the scores above to rank them together.
Rankings by part
We read 8 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Teams, Email and Calendar. Claude Opus 4.6 leads 5 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
Teams
12 models we track have a score here.
- 1 Claude Opus 4.6 52%
- 2 Claude Sonnet 4.5 ↑3 51%
- 3 Claude Opus 4.5 50%
- 4
Gemini 3 Flash ↑2 47.3%
- 5 Claude Sonnet 4.6 ↓3 47%
- 6
Gemini 3.1 Pro ↓2 46%
- 7
Gemini 3 Pro ↑1 43%
- 8
DeepSeek-V3.2 ↑1 35.7%
- 9 Kimi K2 Thinking ↑2 30%
- 10 GPT-5 ↓3 26.3%
12 models we track have a score here.
- 1 Claude Sonnet 4.6 ↑1 58.6%
- 2 Claude Opus 4.6 ↓1 57.7%
- 3 Claude Opus 4.5 51.9%
- 4 Claude Sonnet 4.5 ↑1 51.3%
- 5 Kimi K2 Thinking ↑6 51%
- 6 GPT-5 ↑1 49%
- 7 GPT-5 mini ↑3 47.4%
- 8
Gemini 3.1 Pro ↓4 47.1%
- 9
DeepSeek-V3.2 45.8%
- 10
Gemini 3 Flash ↓4 44.3%
Calendar
12 models we track have a score here.
- 1 Claude Opus 4.6 43.3%
- 2 Claude Opus 4.5 ↑1 43.2%
- 3 GPT-5 ↑4 41.3%
- 4 Claude Sonnet 4.6 ↓2 40.4%
- 5
Gemini 3.1 Pro ↓1 40.4%
- 6 Claude Sonnet 4.5 ↓1 34.6%
- 7
Gemini 3 Flash ↓1 30.5%
- 8 GPT-5 mini ↑2 28.8%
- 9
Gemini 3 Pro ↓1 28.8%
- 10
DeepSeek-V3.2 ↓1 21.5%
HR
12 models we track have a score here.
- 1 Claude Opus 4.6 45.1%
- 2 Claude Sonnet 4.6 37%
- 3 Claude Opus 4.5 32.1%
- 4 Claude Sonnet 4.5 ↑1 21.6%
- 5 GPT-5 ↑2 17.9%
- 6
DeepSeek-V3.2 ↑3 15%
- 7
Gemini 3 Flash ↓1 12.6%
- 8
Gemini 3 Pro 12.5%
- 9
Gemini 3.1 Pro ↓5 10.9%
- 10 GPT-5 mini 10.7%
Drive
12 models we track have a score here.
- 1 Claude Opus 4.6 57.1%
- 2 Claude Sonnet 4.6 57.1%
- 3
Gemini 3.1 Pro ↑1 55.2%
- 4 Claude Sonnet 4.5 ↑1 52.1%
- 5
Gemini 3 Flash ↑1 49.7%
- 6 Claude Opus 4.5 ↓3 49.5%
- 7
Gemini 3 Pro ↑1 46.7%
- 8 Kimi K2 Thinking ↑3 39.6%
- 9 GPT-5 ↓2 34%
- 10
DeepSeek-V3.2 ↓1 27.6%
Hybrid
12 models we track have a score here.
- 1 Claude Opus 4.6 34%
- 2 Claude Opus 4.5 ↑1 30.7%
- 3
Gemini 3.1 Pro ↑1 30.1%
- 4 Claude Sonnet 4.6 ↓2 29.4%
- 5 Claude Sonnet 4.5 28.1%
- 6
Gemini 3 Flash 24.2%
- 7 GPT-5 23.5%
- 8
Gemini 3 Pro 22.9%
- 9
DeepSeek-V3.2 22.9%
- 10 GPT-5 mini 22.5%
Customer service management
12 models we track have a score here.
- 1
Gemini 3.1 Pro ↑3 46.7%
- 2 Claude Opus 4.6 ↓1 45.1%
- 3 GPT-5 ↑4 36.4%
- 4
Gemini 3 Flash ↑2 35%
- 5 Claude Opus 4.5 ↓2 34.2%
- 6 Claude Sonnet 4.6 ↓4 32.6%
- 7
Gemini 3 Pro ↑1 27.7%
- 8 Claude Sonnet 4.5 ↓3 16.7%
- 9 GPT-5 mini ↑1 15.8%
- 10
DeepSeek-V3.2 ↓1 15.4%
IT service management
12 models we track have a score here.
- 1 Claude Sonnet 4.6 ↑1 35.5%
- 2 Claude Opus 4.6 ↓1 33.3%
- 3
Gemini 3.1 Pro ↑1 32.8%
- 4
Gemini 3 Flash ↑2 28.5%
- 5 Claude Opus 4.5 ↓2 23.8%
- 6
Gemini 3 Pro ↑2 22.2%
- 7 GPT-5 18.9%
- 8 Claude Sonnet 4.5 ↓3 17.6%
- 9 Kimi K2 Thinking ↑2 12.2%
- 10
DeepSeek-V3.2 ↓1 9.60%
Number one over time 2 models have held number one with claims included since its release in March 2026
Lab claims count from the day the card came out, independent runs from the model’s release. We draw one row each for the benchmark’s own leaderboard and what labs claim in their cards.
Sources: Official EnterpriseOps-Gym leaderboard and labs’ model cards
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Sources: Official EnterpriseOps-Gym leaderboard and labs’ model cards
3 labs 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 |
|---|---|---|---|
| Nov 20, 2026 | 42 | EnterpriseOps-Gym | |
| Nov 5, 2026 | 46.8 | EnterpriseOps-Gym | |
| Oct 27, 2026 | LFM3-1.7B | 28.2 | EnterpriseOps-Gym |
| Oct 22, 2026 | 33.6 | EnterpriseOps-Gym |
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 |
|---|---|---|---|---|
| Nov 20, 2026 | | 42 | EnterpriseOps-Gym | Capabilities |
Versions
More in Tool use
- AutomationBench In use Agent completes end-to-end business workflows inside a simulated company exposing dozens of REST APIs. Graded on resulting state.
Gemini 4 Neon 56.3%
- MCP Atlas In use Multi-step workflows over real MCP servers. Scores tool-use competency on realistic tasks.
Gemini 4 Argon 90.7%
- BFCL V4 In use Models call functions and tools, including multi-turn agentic tasks with web search and memory. Checked against expected calls and outcomes.
Qwen3.7-Max 75%
- τ³-bench Banking In use Agents act as banking customer-service reps over many turns with a simulated user, following policy. Pass^1 on the final database state.
Qwen 3.8 Max 55.1%
- Toolathlon-Verified In use 108 long-horizon tasks across 32 apps through about 600 MCP tools. Execution-based checkers inspect final environment state. GLM-5.3-Flash 78.4%
- DeepSearchQA In use Questions that need multi-step search and synthesis. Final answers graded against references. Kimi K3 95%
Data audit
All 10 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
- Pass Independent runs Scores from the official leaderboard (15 entries), launch posts (4 scores) and system cards (1 score).
- Pass Models scored 16 models we track, plus 3 other entries on its leaderboard.
- Pass Newest score The newest is from November 2026, for Inkling 2 Mini.
- Note Frontier models None of the top 10 has a score, which is fine for a retired test. The strongest tested is Claude Opus 4.6 (#39).
- Pass Headroom The best is 46.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on github.com: 12 rows for models we track and 3 other entries at the last read.
- Pass Last read November 18th, 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.