Z.ai · open weights on Arena · first on Arena 10 February 2026

glm-5

88 days for open weights to match gemini-3-pro

This open-weights model matched the score of gemini-3-pro in WebDev 88 days after gemini-3-pro took first place.

Embed

Credits LMArena.

Credits LMArena.

Credits Epoch AI.

Each frame links back to its page. Preview it.

Cite
Download
Post on X

Text: best place 7

place 61 of 419 on 28 October 2026
1st10th
On the board
10 February 2026 to 28 October 2026, the newest snapshot
Score
1446 (1442–1450) on 28 October 2026 · 29,255 votes

WebDev: best place 6

place 87 of 152 on 28 October 2026
1st10th
On the board
12 February 2026 to 28 October 2026, the newest snapshot
Score
1436 (1427–1444) on 28 October 2026 · 7,465 votes
Matched
  • gemini-3-pro on 12 February 2026, 88 days after it took first place

Epoch AI: GLM-5

released 11 February 2026 · Open weights (unrestricted)
  1. FrontierMath-2025-02-28-Private 16.4%
  2. FrontierMath-Tier-4-2025-07-01-Private 2.1%
  3. GPQA diamond 87.8%
  4. OTIS Mock AIME 2024-2025 80.0%
  5. SWE-Bench verified 72.1%

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.