Benchmark Almanac

SArena

InternSVG 8B 0.95 tops its leaderboard Claude Sonnet 4 0.92 leads the models we track
Intern-S1-Pro 83.5 leads since February 2026
Fading

Models understand, edit and generate SVG graphics in domains such as icons, illustrations and chemistry diagrams. Generation is scored by comparing rendered output with references.

Released with the InternSVG model and the SAgoge dataset. Labs report the Icon subset.

Released
October 2025
Kind
Web and app front ends
Level
Everyday
Scored
Checked automatically
Answers
Code
Tasks
All public
Strongest model tested
Claude Sonnet 4 #112 on the capability index
Last in a launch post
February 2026, dropped before it saturated
Last model added
March 2026
Scores refreshed
November 12th, 2026
26
Entries on its own leaderboard, 17 models we track
1
Launch post lists it, from 1 lab
2
System cards report it, from 1 lab

InternSVG 8B tops the leaderboard with 0.95

Best score per entry from the benchmark's own leaderboard, read on November 12th, 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. The best model we track is Claude Sonnet 4, at 0.92.

Intern-S1-Pro leads with 83.5, 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.

The newest model with a score here came out in August 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

26 entries. Scroll sideways for the rest.

  1. 0.95 InternSVG 8B
  2. 0.92 Claude Sonnet 4
  3. 0.91 Claude-Sonnet-3.7
  4. 0.89 OmniSVG 3B
  5. 0.88 Gemini-2.5-Flash (snapshot unspecified)
  6. 0.87 GPT-4o (snapshot unspecified)
  7. 0.87 GLM-4.5V
  8. 0.87 Starvector 8B
  9. 0.86 Llama 4 Maverick
  10. 0.85 InternVL3-78B
  11. 0.84 Llama 4 Scout
  12. 0.84 Qwen2.5-VL-72B
  13. 0.84 Qwen2.5-VL-32B
  14. 0.83 Step3-321B
  15. 0.83 InternVL3-38B
  16. 0.83 Gemma 3 27B
  17. 0.82 InternVL3-14B
  18. 0.82 Gemma 3 12B
  19. 0.82 GLM-4.1V-9B-Thinking
  20. 0.81 InternVL3-8B
  21. 0.80 Keye-VL-8B-Preview
  22. 0.80 Kimi-VL-A3B
  23. 0.78 Qwen2.5-VL-7B
  24. 0.76 Llama 3.2 11B
  25. 0.76 Llama 3.2 90B
  26. 0.75 LLM4SVG 7B
  1. 1 InternSVG 8B SArena-Icon; image-to-SVG DINO 0.95
  2. 2 Claude Sonnet 4 SArena-Icon; image-to-SVG DINO 0.92
  3. 3 Claude-Sonnet-3.7 SArena-Icon; image-to-SVG DINO 0.91
  4. 4 OmniSVG 3B SArena-Icon; image-to-SVG DINO 0.89
  5. 5 Gemini-2.5-Flash (snapshot unspecified) SArena-Icon; image-to-SVG DINO 0.88
  6. 6 GPT-4o (snapshot unspecified) SArena-Icon; image-to-SVG DINO 0.87
  7. 7 GLM-4.5V SArena-Icon; image-to-SVG DINO 0.87
  8. 8 Starvector 8B SArena-Icon; image-to-SVG DINO 0.87
  9. 9 Llama 4 Maverick SArena-Icon; image-to-SVG DINO 0.86
  10. 10 InternVL3-78B SArena-Icon; image-to-SVG DINO 0.85
  11. 11 Llama 4 Scout SArena-Icon; image-to-SVG DINO 0.84
  12. 12 Qwen2.5-VL-72B SArena-Icon; image-to-SVG DINO 0.84
  13. 13 Qwen2.5-VL-32B SArena-Icon; image-to-SVG DINO 0.84
  14. 14 Step3-321B SArena-Icon; image-to-SVG DINO 0.83
  15. 15 InternVL3-38B SArena-Icon; image-to-SVG DINO 0.83
  16. 16 Gemma 3 27B SArena-Icon; image-to-SVG DINO 0.83
  17. 17 InternVL3-14B SArena-Icon; image-to-SVG DINO 0.82
  18. 18 Gemma 3 12B SArena-Icon; image-to-SVG DINO 0.82
  19. 19 GLM-4.1V-9B-Thinking SArena-Icon; image-to-SVG DINO 0.82
  20. 20 InternVL3-8B SArena-Icon; image-to-SVG DINO 0.81
  21. 21 Keye-VL-8B-Preview SArena-Icon; image-to-SVG DINO 0.80
  22. 22 Kimi-VL-A3B SArena-Icon; image-to-SVG DINO 0.80
  23. 23 Qwen2.5-VL-7B SArena-Icon; image-to-SVG DINO 0.78
  24. 24 Llama 3.2 11B SArena-Icon; image-to-SVG DINO 0.76
  25. 25 Llama 3.2 90B SArena-Icon; image-to-SVG DINO 0.76
  26. 26 LLM4SVG 7B SArena-Icon; image-to-SVG DINO 0.75

Source: Official SArena leaderboard

18 models. Scroll sideways for the rest.

  1. 83.5 Intern-S1-Pro
  2. 0.92 Claude Sonnet 4
  3. 0.87 GLM-4.5V
  4. 0.86 Llama 4 Maverick
  5. 0.85 InternVL3-78B
  6. 0.84 Llama 4 Scout
  7. 0.84 Qwen2.5-VL-72B
  8. 0.84 Qwen2.5-VL-32B
  9. 0.83 InternVL3-38B
  10. 0.83 Gemma 3 27B
  11. 0.82 InternVL3-14B
  12. 0.82 Gemma 3 12B
  13. 0.82 GLM-4.1V-9B-Thinking
  14. 0.81 InternVL3-8B
  15. 0.80 Keye-VL-8B-Preview
  16. 0.78 Qwen2.5-VL-7B
  17. 0.76 Llama 3.2 11B
  18. 0.76 Llama 3.2 90B
  1. 1 Intern-S1-Pro 83.5
  2. 2 Claude Sonnet 4 SArena-Icon; image-to-SVG DINO 0.92
  3. 3 GLM-4.5V SArena-Icon; image-to-SVG DINO 0.87
  4. 4 Llama 4 Maverick SArena-Icon; image-to-SVG DINO 0.86
  5. 5 InternVL3-78B SArena-Icon; image-to-SVG DINO 0.85
  6. 6 Llama 4 Scout SArena-Icon; image-to-SVG DINO 0.84
  7. 7 Qwen2.5-VL-72B SArena-Icon; image-to-SVG DINO 0.84
  8. 8 Qwen2.5-VL-32B SArena-Icon; image-to-SVG DINO 0.84
  9. 9 InternVL3-38B SArena-Icon; image-to-SVG DINO 0.83
  10. 10 Gemma 3 27B SArena-Icon; image-to-SVG DINO 0.83
  11. 11 InternVL3-14B SArena-Icon; image-to-SVG DINO 0.82
  12. 12 Gemma 3 12B SArena-Icon; image-to-SVG DINO 0.82
  13. 13 GLM-4.1V-9B-Thinking SArena-Icon; image-to-SVG DINO 0.82
  14. 14 InternVL3-8B SArena-Icon; image-to-SVG DINO 0.81
  15. 15 Keye-VL-8B-Preview SArena-Icon; image-to-SVG DINO 0.80
  16. 16 Qwen2.5-VL-7B SArena-Icon; image-to-SVG DINO 0.78
  17. 17 Llama 3.2 11B SArena-Icon; image-to-SVG DINO 0.76
  18. 18 Llama 3.2 90B SArena-Icon; image-to-SVG DINO 0.76

Sources: Labs’ model cards and official SArena leaderboard

#EntryScoreSettingDateSource
1 InternSVG 8B 0.95 SArena-Icon; image-to-SVG DINO Official leaderboard
2 Claude Sonnet 4 0.92 SArena-Icon; image-to-SVG DINO May 2025 Official leaderboard
3 Claude-Sonnet-3.7 0.91 SArena-Icon; image-to-SVG DINO Official leaderboard
4 OmniSVG 3B 0.89 SArena-Icon; image-to-SVG DINO Official leaderboard
5 Gemini-2.5-Flash (snapshot unspecified) 0.88 SArena-Icon; image-to-SVG DINO Official leaderboard
6 GPT-4o (snapshot unspecified) 0.87 SArena-Icon; image-to-SVG DINO Official leaderboard
7 GLM-4.5V 0.87 SArena-Icon; image-to-SVG DINO Aug 2025 Official leaderboard
8 Starvector 8B 0.87 SArena-Icon; image-to-SVG DINO Official leaderboard
9 Llama 4 Maverick 0.86 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
10 InternVL3-78B 0.85 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
11 Llama 4 Scout 0.84 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
12 Qwen2.5-VL-72B 0.84 SArena-Icon; image-to-SVG DINO Jan 2025 Official leaderboard
13 Qwen2.5-VL-32B 0.84 SArena-Icon; image-to-SVG DINO Mar 2025 Official leaderboard
14 Step3-321B 0.83 SArena-Icon; image-to-SVG DINO Official leaderboard
15 InternVL3-38B 0.83 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
16 Gemma 3 27B 0.83 SArena-Icon; image-to-SVG DINO Mar 2025 Official leaderboard
17 InternVL3-14B 0.82 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
18 Gemma 3 12B 0.82 SArena-Icon; image-to-SVG DINO Mar 2025 Official leaderboard
19 GLM-4.1V-9B-Thinking 0.82 SArena-Icon; image-to-SVG DINO Jul 2025 Official leaderboard
20 InternVL3-8B 0.81 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
21 Keye-VL-8B-Preview 0.80 SArena-Icon; image-to-SVG DINO Jun 2025 Official leaderboard
22 Kimi-VL-A3B 0.80 SArena-Icon; image-to-SVG DINO Official leaderboard
23 Qwen2.5-VL-7B 0.78 SArena-Icon; image-to-SVG DINO Jan 2025 Official leaderboard
24 Llama 3.2 11B 0.76 SArena-Icon; image-to-SVG DINO Sep 2024 Official leaderboard
25 Llama 3.2 90B 0.76 SArena-Icon; image-to-SVG DINO Sep 2024 Official leaderboard
26 LLM4SVG 7B 0.75 SArena-Icon; image-to-SVG DINO Official leaderboard
#ModelScoreSettingDateSource
1 Intern-S1-Pro 83.5 Feb 2026 Lab model card
2 Claude Sonnet 4 0.92 SArena-Icon; image-to-SVG DINO May 2025 Official leaderboard
3 GLM-4.5V 0.87 SArena-Icon; image-to-SVG DINO Aug 2025 Official leaderboard
4 Llama 4 Maverick 0.86 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
5 InternVL3-78B 0.85 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
6 Llama 4 Scout 0.84 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
7 Qwen2.5-VL-72B 0.84 SArena-Icon; image-to-SVG DINO Jan 2025 Official leaderboard
8 Qwen2.5-VL-32B 0.84 SArena-Icon; image-to-SVG DINO Mar 2025 Official leaderboard
9 InternVL3-38B 0.83 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
10 Gemma 3 27B 0.83 SArena-Icon; image-to-SVG DINO Mar 2025 Official leaderboard
11 InternVL3-14B 0.82 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
12 Gemma 3 12B 0.82 SArena-Icon; image-to-SVG DINO Mar 2025 Official leaderboard
13 GLM-4.1V-9B-Thinking 0.82 SArena-Icon; image-to-SVG DINO Jul 2025 Official leaderboard
14 InternVL3-8B 0.81 SArena-Icon; image-to-SVG DINO Apr 2025 Official leaderboard
15 Keye-VL-8B-Preview 0.80 SArena-Icon; image-to-SVG DINO Jun 2025 Official leaderboard
16 Qwen2.5-VL-7B 0.78 SArena-Icon; image-to-SVG DINO Jan 2025 Official leaderboard
17 Llama 3.2 11B 0.76 SArena-Icon; image-to-SVG DINO Sep 2024 Official leaderboard
18 Llama 3.2 90B 0.76 SArena-Icon; image-to-SVG DINO Sep 2024 Official leaderboard

Labs report scores for 1 model in their cards. Switch the scores above to rank them together.

Number one over time 2 models have held number one with claims included since its release in October 2025

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Own leaderboard Not tested since Oct 2025
Claude Sonnet 4 0.92

Source: Official SArena leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Reported by the lab in its model card
0.00 20.0 40.0 60.0 80.0 100.0 2025 2026 0.00 20.0 40.0 60.0 80.0 100.0 2025 2026

Sources: Official SArena leaderboard and labs’ model cards

Labs headline it less often now, last in February 2026

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.

Shanghai AI Lab
DateLaunch postScoreAs the post writes it
Feb 2, 2026 Intern-S1-Pro 83.5 SArena (Icon)

2 system cards report it, from 1 lab

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.

DateDocumentScoreAs the document writes itSection
Mar 26, 2026 Intern-S1-Pro technical report 83.5 SArena (Icon), SVG Generation, thinking (Table 2) Capabilities
Aug 25, 2025 InternVL3.5-241B-A28B technical report SArena-Icon Text2SVG FID 11.27, FID-C 4.43, CLIP 76.81. Img2SVG DINO 0.88, SSIM 0.64, LPIPS 0.29, PSNR 9.19, no overall score (Table 13) Capabilities

More in Coding

All 282 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471