Price from Alibaba: the cheapest offer that is serving, at standard delivery and at the maker’s precision or better, at its standard rate.
Every price dimension · USD per million tokens
Input
$2.00/M
Output
$6.00/M
Cached input
$0.250/M
Cache write
$2.50/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 $2.00/M from SiliconFlow.
7 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)
Seller
SiliconFlowsiliconflow/fp8 Endpoint terms
Cached input /M
$0.250
Context limit
1,048,576 tokens
Output limit
131,072 tokens
Tools
Listed by endpoint
Reasoning
Listed by endpoint
Promotional discount
None reported
Uptime · last 30 minutes
99.80%
Input$2.00
Output$6.00
Precisionfp8
Uptime (1d)99.53%
Seller
DeepInfradeepinfra/fp4 Endpoint terms
Cached input /M
$0.200
Context limit
262,144 tokens
Output limit
131,072 tokens
Tools
Listed by endpoint
Reasoning
Listed by endpoint
Promotional discount
None reported
Uptime · last 30 minutes
100.00%
Input$2.00
Output$6.00
Precisionfp4
Uptime (1d)99.81%
Seller
Novitanovita Endpoint terms
Cached input /M
$0.250
Context limit
1,000,000 tokens
Output limit
131,072 tokens
Tools
Listed by endpoint
Reasoning
Listed by endpoint
Promotional discount
None reported
Uptime · last 30 minutes
100.00%
Input$2.00
Output$6.00
Precisionundeclared
Uptime (1d)99.90%
Seller
Alibabaalibaba Endpoint terms
Cached input /M
$0.250
Context limit
1,000,000 tokens
Output limit
131,072 tokens
Tools
Listed by endpoint
Reasoning
Listed by endpoint
Promotional discount
None reported
Uptime · last 30 minutes
100.00%
Input$2.00
Output$6.00
Precisionundeclared
Uptime (1d)99.90%
Seller
Venicevenice Endpoint terms
Cached input /M
$0.250
Context limit
262,144 tokens
Output limit
65,536 tokens
Tools
Listed by endpoint
Reasoning
Listed by endpoint
Promotional discount
20% · already included in these rates
Uptime · last 30 minutes
Unknown
Input$2.00
Output$6.00
Precisionundeclared
Uptime (1d)99.77%
Seller
Modalmodal Endpoint terms
Cached input /M
$0.250
Context limit
1,000,000 tokens
Output limit
262,144 tokens
Tools
Listed by endpoint
Reasoning
Listed by endpoint
Promotional discount
None reported
Uptime · last 30 minutes
100.00%
Input$2.00
Output$6.00
Precisionundeclared
Uptime (1d)98.91%
Seller
Togethertogether Endpoint terms
Cached input /M
$0.250
Context limit
1,010,000 tokens
Output limit
909,000 tokens
Tools
Listed by endpoint
Reasoning
Listed by endpoint
Promotional discount
None reported
Uptime · last 30 minutes
100.00%
Input$2.00
Output$6.00
Precisionundeclared
Uptime (1d)99.96%
Independent scores
No independent benchmark has measured this model. Unrated is not a score of zero.
Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of [Qwen3.8 Max](/qwen/qwen3.8-max), with 95 billion active parameters out of 2.4 trillion total. It is...
Questions this page answers
What does Qwen3.8 2.4T A95B cost?
$2.00/M in, $6.00/M out in this catalogue, as of the fetch date on this page. That is the API row, not a subscription.
Does Qwen3.8 2.4T A95B have an independent quality score?
Qwen3.8 2.4T A95B has no independent quality score in this catalogue. Unrated is not a score of zero.
What beats Qwen3.8 2.4T A95B?
Qwen3.8 2.4T A95B has no independent quality score. Unrated is not a score of zero.
Does this page use Artificial Analysis scores for Qwen3.8 2.4T A95B?
No. Artificial Analysis figures are not published here. Quality on this page is LMArena Elo where a score exists; otherwise the row is unrated.
Does a missing score mean Qwen3.8 2.4T A95B scored zero?
Unrated is not a score of zero.
Is the cheapest Qwen3.8 2.4T A95B endpoint the same product?
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $2.00/M from SiliconFlow.
MODEL MONUMENT
Save or share this proof
Preview and more sharing options
Qwen: Qwen3.8 2.4T A95B. Balanced workload. Not independently rated. Evidence as of 06 Oct 2026.
Related
No independent quality score, so there is no dominator to name. Unrated is not a clean bill of health.
The verdict as one SVG, for a README or a docs page. It states this model's dominance
status at the balanced workload on the LMArena lens, with the date it was computed, and
it is rebuilt with the catalogue — so it changes when the verdict changes, including to
one you would rather it did not.