DeepSeek · open weights on Arena · first on Arena 30 December 2024

deepseek-v3

229 days for open weights to match gpt-4o-2024-05-13

This open-weights model matched the score of gpt-4o-2024-05-13 in Text 229 days after gpt-4o-2024-05-13 took first place. It also matched grok-2-2024-08-13 in Text.

Embed

Who held first place

<iframe src="https://jesseahawkins.com/frontier/embed/" title="Frontier: who held first place on each Arena leaderboard, and for how long" width="100%" height="560" style="border:0;max-width:100%" loading="lazy"></iframe>

Credits LMArena.

The open-weights catch-up

<iframe src="https://jesseahawkins.com/frontier/embed/catch-up/" title="Frontier: the days until open weights matched each closed leader" width="100%" height="900" style="border:0;max-width:100%" loading="lazy"></iframe>

Credits LMArena.

The saturation clock

<iframe src="https://jesseahawkins.com/frontier/embed/clock/" title="Frontier: the days from each benchmark’s release to a model at 90%" width="100%" height="640" style="border:0;max-width:100%" loading="lazy"></iframe>

Credits Epoch AI.

Each frame links back to its page. Preview it.

Cite

This page

Jesse Hawkins, “deepseek-v3 · Frontier”, jesseahawkins.com, data retrieved 30 Oct 2026. https://jesseahawkins.com/frontier/deepseek-v3/. Data: LMArena, Arena leaderboard dataset, revision 4491b06, CC BY 4.0; Epoch AI, Capabilities & benchmarking, CC BY 4.0.

BibTeX

@misc{hawkins2026frontierdeepseekv3,
  author = {Hawkins, Jesse},
  title = {{deepseek-v3 · Frontier}},
  year = {2026},
  howpublished = {\url{https://jesseahawkins.com/frontier/deepseek-v3/}},
  note = {Data: LMArena, Arena leaderboard dataset, revision 4491b06, CC BY 4.0; Epoch AI, Capabilities \& benchmarking, CC BY 4.0, retrieved 2026-10-30}
}

Arena · LMArena

LMArena, "Arena leaderboard dataset", Hugging Face, revision 4491b06. https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset. CC BY 4.0.

Benchmarks · Epoch AI

Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks' [online resource]. CC BY 4.0.

Download

Post on X

Text: best place 3

place 227 of 419 on 28 October 2026

1st10th100th

On the board: 30 December 2024 to 28 October 2026, the newest snapshot
Score: 1332 (1328–1337) on 28 October 2026 · 21,770 votes
Matched: gpt-4o-2024-05-13 on 30 December 2024, 229 days after it took first place
grok-2-2024-08-13 on 30 December 2024, 106 days after it took first place

Epoch AI: DeepSeek-V3

released 26 December 2024 · Open weights (restricted use)

FrontierMath-2025-02-28-Private 1.7%
GPQA diamond 56.5%
MATH level 5 64.9%
OTIS Mock AIME 2024-2025 15.8%

Sources

Arena scores come from people voting between two anonymous answers. First place is the highest score; the top two are often within each other’s margin of error. LMArena, Arena leaderboard dataset, revision 4491b06, retrieved 30 October 2026; Epoch AI, Capabilities & benchmarking, retrieved 30 October 2026. Both CC BY 4.0.