AI model market · 2026-44
DeepSeek V4.1 Flash 1017 rose 15.5 points to 17.8% of OpenRouter tokens
As of 26 October 2026, DeepSeek V4.1 Flash 1017 was the biggest riser in OpenRouter usage, up 15.5 points to a 17.8% share of tokens over the seven days ending 25 October 2026. Gemini 4 Argon led the LMArena text leaderboard at 1,525.7. OpenRouter listed 35 price changes over the same 7-day period.
Produced automatically from OpenRouter and LMArena data.
- Most-used model
- DeepSeek V4.1 Flash 1017
- Biggest riser
- DeepSeek V4.1 Flash 1017
- Cheapest frontier model
- MiMo-V2.6-Pro
- New models
- 1
17.8%
+15.5
$0.435 / $0.87
What changed
- DeepSeek V4.1 Flash fell 10.4 points to a 9.6% share of OpenRouter tokens, while Step 5 Preview fell 5.1 points to 1.9% in the same period.
- GPT-6 Luna and Claude Opus 5.5 each held a 4.4% share of tokens, and each rose 0.6 points, ranking sixth and seventh among the ten models listed.
- Gemini 4 Argon held first place on the LMArena text leaderboard with a score of 1,525.7, ahead of Claude Opus 5.5 at 1,516.2 in second place.
- Muse Spark 1.3 moved up 2 places to arena rank 10 with 1,491.1, while Claude Fable 5 moved down 2 places to rank 12 with a score of 1,490.9.
- Among 35 listed price changes, DeepSeek V4 Flash 0731 showed the largest rise in output price, from $0.15 to $1.32, up 780.0%, followed by DeepSeek V4.1 Flash at $1.00, up 733.3%.
Why it matters
Rankings and usage shares show builders which models are carrying the most traffic on OpenRouter. DeepSeek V4.1 Flash 1017 took 17.8% of OpenRouter tokens, while DeepSeek V4.1 Flash, at 9.6%, fell 10.4 points. Shares across the 369 tracked models can move by double digits within a single 7-day period, as these two listings show. Both movements were measured over the 7 days ending 25 October 2026. The usage data comes from OpenRouter rankings.
Price and quality data matter to buyers as well. MiMo-V2.6-Pro was the cheapest frontier model among the LMArena top 15, at $0.435 input and $0.87 output per million tokens, and it sat at arena rank 11. Gemma 4 26B A4B led best value, at 13,418.5 arena score per blended dollar. OpenRouter also added one new model in the past 7 days, Kimi K3.1 Pro, priced at $2.00 input and $10.00 output per million tokens with 1.05M tokens of context.
The numbers
| Rank | Model | Share, 7 days | Change (points) | Input $/1M | Output $/1M |
|---|---|---|---|---|---|
| 1 | DeepSeek V4.1 Flash 1017 | 17.8% | +15.5 | $0.30 | $1.20 |
| 2 | DeepSeek V4.1 Flash | 9.6% | -10.4 | $0.02 | $1.00 |
| 4 | GLM 5.3 Flash | 7.0% | -0.5 | $0.15 | $0.50 |
| 5 | Hy4 preview | 4.8% | +0.2 | $0.834 | $2.501 |
| 6 | GPT-6 Luna | 4.4% | +0.6 | $0.10 | $0.50 |
| 7 | Claude Opus 5.5 | 4.4% | +0.6 | $4.00 | $20.00 |
| 8 | MiMo-V2.6-Flash | 3.2% | -2.9 | $0.14 | $0.28 |
| 10 | GLM 5.3 | 2.7% | -0.5 | $0.0699 | $1.40 |
| 11 | DeepSeek V4 Flash 0731 | 2.1% | -0.8 | $0.44 | $1.32 |
| 12 | Step 5 Preview | 1.9% | -5.1 | $1.00 | $2.70 |
Methodology and sources
Written automatically by a language model from daily snapshots of the OpenRouter model catalogue, the OpenRouter rankings (CC BY 4.0) and the LMArena text leaderboard (CC BY 4.0). Snapshot date 26 October 2026; comparison: 7-day change. The model was given the numbers only and told to add nothing else; numbers and wording are checked by program before publication. Live figures: /ai-models.
Cite this page
Oesnada, "DeepSeek V4.1 Flash 1017 rose 15.5 points to 17.8% of OpenRouter tokens", https://oesnada.com/analysis/ai-model-market/2026-44, accessed 26 October 2026.