MODEL PROOF
Qwen3.5 397B A17B
BETTER VALUE OPTION
A stored alternative has an equal-or-higher measured score and an equal-or-lower price.
Serving precision differs between offers.
MiMo-V2.6-Pro · Δ score 55.1 points · 38% lower measured price · $0.540 / 1M tokens
DeepSeek V4.1 Flash is a further stored option with these losses: no video input
- Independent LMArena score
- 1,437.6 ± 3.03 · 97,351 votes
- Context window
- 262.14K
- Maximum output
- 81.92K
- Input modalities
- text, image, video
- Output modalities
- text
- Published input price
- $0.390 / 1M tokens
- Published output price
- $2.34 / 1M tokens
- Pricing kind
- fixed
Inspect complete billing conditions and endpoint terms below.
Price from Alibaba: the cheapest offer that is serving, at standard delivery, with at least 90% of the reference context and at the maker’s precision or better, at its standard rate.
Cheaper right now: $0.870/M at DeepInfra — delivery tier (flex) It does not pass the like-for-like test, so it never sets a rank.
Every price dimension · USD per million tokens
| Input | $0.390/M |
|---|---|
| Output | $2.34/M |
| Cached input | Unknown/M |
| Cache write | Unknown/M |
| Cache write 1h | Unknown/M |
| Reasoning | Unknown/M |
Who sells it
The same weights, different shops. Cheapest is not like-for-like when serving precision differs.
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $0.450/M from DeepInfra.
8 provider offers · rates, limits and conditions
Rates are USD per million tokens. Endpoint terms can differ even when the quoted price and precision match. A listed parameter is a provider declaration, not a task-success test.
| Seller | Input | Output | Precision | Uptime (1d) |
|---|---|---|---|---|
| DeepInfradeepinfra/fp8/flex Endpoint terms
| $0.360 | $2.40 | fp8 | 100.00% |
| Alibabaalibaba Endpoint terms
| $0.390 | $2.34 | undeclared | 99.56% |
| DeepInfradeepinfra/fp8 Endpoint terms
| $0.450 | $3.00 | fp8 | 100.00% |
| Parasailparasail/fp8 Endpoint terms
| $0.500 | $3.60 | fp8 | 100.00% |
| StreamLakestreamlake Endpoint terms
| $0.600 | $3.60 | undeclared | 98.67% |
| Novitanovita Endpoint terms
| $0.600 | $3.60 | undeclared | 100.00% |
| DeepInfradeepinfra/fp8/priority Endpoint terms
| $0.675 | $4.50 | fp8 | 98.74% |
| Venicevenice Endpoint terms
| $0.750 | $4.50 | undeclared | 99.23% |
Independent scores
| LMArena Elo | 1437.6default |
|---|---|
| LMArena row | qwen3.5-397b-a17b |
LMArena Elo under CC BY 4.0. Absence of another board is not a score of zero.
Best rankings by task
- SVG #62 1123
- Data visualisation #72 1170
- Code categories #76 1161
- Websites #77 1178
- 3D #78 1128
- UI components #86 1147
- Game development #88 1112
Per-task Elo and rank from Design Arena, via OpenRouter’s model feed. The rank is the feed’s, and it counts models OpenRouter does not list.
Capability
| Context window | 262K |
|---|---|
| Max output | 81K |
| Input modes | text, image, video |
| Tool use | yes |
| Reasoning | optional |
| Open weights | Unknown |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-11-22 |
| Quality data | verified |
The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers...
Questions this page answers
What does Qwen3.5 397B A17B cost?
$0.390/M in, $2.34/M out in this catalogue, as of the fetch date on this page. That is the API row, not a subscription.
Does Qwen3.5 397B A17B have an independent quality score?
Qwen3.5 397B A17B has an LMArena score in this catalogue.
What beats Qwen3.5 397B A17B?
MiMo-V2.6-Pro has an equal-or-higher measured score and an equal-or-lower price.
Does this page use Artificial Analysis scores for Qwen3.5 397B A17B?
No. Artificial Analysis figures are not published here. Quality on this page is LMArena Elo where a score exists; otherwise the row is unrated.
Is the cheapest Qwen3.5 397B A17B endpoint the same product?
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $0.450/M from DeepInfra.
MODEL MONUMENT