DeepSeek-V4.1-Flash

Leaderboards

Where DeepSeek-V4.1-Flash sits among the top 20 in each category.

By task

Scored from votes on each kind of prompt alone, on the same 1,000-point scale.

Test results

Every test DeepSeek-V4.1-Flash has a published result on, out of 100%, with its rank among the models that took it.

• GPQA

  #23 / 256

  Graduate-level science questions written to be hard to look up.

  90.9%

  Self-reported

• Humanity's Last Exam

  #44 / 108

  Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions

  36.8%

  Self-reported · In the text score· In the vision score

• DeepSWE 1.1

  #10 / 50

  DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.

  74.2%

  Self-reported · In the coding score

• Terminal-Bench 2.1

  #1 / 46

  Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.

  90.6%

  Self-reported

• Terminal-Bench 4.0

  #18 / 28

  Terminal-Bench 4.0 evaluates AI agents on real-world work in terminal and command-line containerized environments across 66 tasks, with emphasis on science-adjacent and frontier engineering problems. Relative to earlier Terminal-Bench releases, 4.0 increases timeouts and adaptively raises RAM/CPU on selected tasks to reduce harness and resource confounds.

  31.2%

  Self-reported · In the coding score· In the agents score

• NL2Repo

  #2 / 25

  NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.

  64.0%

  Self-reported · In the coding score

• Agents' Last Exam

  #14 / 24

  Agents' Last Exam is a challenging benchmark for AI agents on hard, long-horizon tasks that test sustained reasoning, planning, and tool use, reported with and without tool access.

  31.8%

  Self-reported · In the agents score

• AutomationBench

  #3 / 24

  AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.

  54.8%

  Self-reported · In the agents score

• CyberGym

  #3 / 21

  CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.

  88.1%

  Self-reported

• CodeForces

  #1 / 20

  A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.

  115.7%

  Self-reported

• Program Bench

  #16 / 18

  Program Bench evaluates code-generation agents by asking them to recreate a program's behavior from only a compiled binary and documentation. It spans 200 tasks from small CLI tools to large systems such as FFmpeg and SQLite, with submissions judged against more than 248,000 fuzz-generated behavioral tests.

  20.3%

  Self-reported · In the coding score

• BabyVision

  #3 / 16

  A benchmark for early-stage visual reasoning and perception on child-like vision tasks.

  89.6%

  Self-reported

• ZEROBench

  #4 / 15

  ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.

  0.5%

  Self-reported

• Humanity's Last Exam (no tools, text-only)

  #10 / 14

  Text-only Humanity's Last Exam variant evaluated without tool use.

  39.1%

  Self-reported

• MathArena Apex

  #5 / 11

  MathArena Apex is a challenging math contest benchmark featuring the most difficult mathematical problems designed to test advanced reasoning and problem-solving abilities of AI models. It focuses on olympiad-level mathematics and complex multi-step mathematical reasoning.

  65.6%

  Self-reported

• ExploitGym

  #5 / 8

  ExploitGym is a large-scale, realistic benchmark built from real-world vulnerabilities across userspace programs, Google's V8 engine, and the Linux kernel. Given a vulnerability and a proof-of-vulnerability input, agents must craft a working end-to-end exploit that achieves unauthorized code execution.

  15.3%

  Self-reported

• SEC-bench Pro

  #4 / 7

  SEC-bench Pro is a self-evolving software-security benchmark that measures agent bug hunting on critical, high-complexity systems. It instantiates validated vulnerabilities across the V8 and SpiderMonkey JavaScript engines as reproducible vulnerability-discovery and proof-of-concept-generation tasks with oracle-based validation.

  62.8%

  Self-reported

• Chartography

  #5 / 6

  Chartography evaluates professional chart understanding with 100 expert-written tasks drawn from real charts across science, engineering, finance, manufacturing, supply chain, and healthcare.

  78.9%

  Self-reported

• Humanity's Last Exam (with tools)

  #2 / 4

  Full multimodal Humanity's Last Exam evaluated with tool use enabled, distinct from its text-only subset.

  63.9%

  Self-reported

• Terminal-Bench 3.0

  #1 / 4

  Terminal-Bench 3.0 is a release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal on real-world, end-to-end tasks.

  30.0%

  Self-reported