Pick models to compare
Click a model to add it, then ask your question below.
Intelligence: the Artificial Analysis Intelligence Index at each model's highest effort. Models without a score yet sit in the lane under the chart, placed by price. Lines at $2 blended and the median score (39.5).
Nemotron 3 UltravsQwen3.8 2.4T A95B
Send one prompt to every model and read the answers side by side, with speed and cost for each.
Shareable link:
Before you run
Highlighted values are the best among your picks.
Nemotron 3 UltraNVIDIA· open weights Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy | Qwen3.8 2.4T A95BAlibaba· open weights Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows | |
|---|---|---|
| Intelligence | ||
| Intelligence index | 22.9 | 39.9 |
| Price | ||
| Input, $ per 1M | $0.60 | $2 |
| Output, $ per 1M | $2.40 | $6 |
| Typical message (500 in / 250 out) | $0.00090 | $0.0025 |
| Repeated input, cached | $0.12 · 5× cheaper | $0.25 · 8× cheaper |
| Prompts over 200K tokens | Same price | Same price |
| Limits | ||
| Context window | 1M | 262K |
| Longest answer | 65K tokens | 128K tokens |
| Knowledge cutoff | Not published | Not published |
| What it can take and do | ||
| Accepts | Text | TextImage |
| Reasoning effort | Off · Low · Med · High | Low · Med · X-high |
| Tool use | Yes | Yes |
| Guaranteed JSON output | Not listed | Yes |
| Availability | ||
| Released | 4 Jun 2026 | 12 Aug 2026 |
| Open weights | Yes, can self-host | Yes, can self-host |
| API providers offering it | 23 | 23 |