Qwen 3.8 Flash is Alibaba's cheapest model at $0.16/1M input tokens — 94% less than the flagship Qwen 3.7 Max. It is also the best capability-per-dollar pick in the lineup.
The shortest way to see the safest default, the lower-cost option, and the specialist pick before you read deeper.
Qwen 3.8 Flash is the safest overall answer here when you want the strongest default instead of the lowest list price.
Mistral: Mistral Nemo is the lower-cost option to start with when you still need useful output at scale.
Qwen 3.8 Max is the better pick when response speed matters more than maximum reasoning depth.
Qwen 3.8 Flash is the lowest-cost Alibaba model: $0.16/1M input, $0.47/1M output.
Qwen 3.8 Flash is the best capability-per-dollar pick (budget score 93/100).
Qwen 3.7 Max costs 16x more on input — reserve it for work where quality is the bottleneck.
Choose Qwen 3.8 Flash for high-volume, low-stakes tasks like classification, extraction, and drafts.
Choose Qwen 3.8 Flash as the everyday default if you want one budget model.
Route only the hardest tasks to Qwen 3.7 Max — a two-tier setup usually cuts spend 60–80%.
Switch the scoring lens to see whether the top answer changes when you care more about cost, speed, or long-document work.
Alibaba / Balanced / Aug 6, 2026
Best Chinese flagship — beats GPT-5.6 Sol on coding, #2 globally for vision.
Ranks models by the broadest mix of coding, writing, research, and long-context usefulness.
You need independently verified benchmarks or Western data residency.
The fastest way to see where the recommendation shifts when your priority changes.
SWE-bench Pro 62.5 at sixteen cents per million input.
Best Chinese flagship — beats GPT-5.6 Sol on coding, #2 globally for vision.
Agent-first Qwen flagship, superseded by Qwen 3.8 Max.
Every figure below is the provider's list price or a published capability score — the same numbers the recommendation on this page is built from.
| Model | Input | Output | Est. month | Context | Speed | Coding | Writing | Research |
|---|---|---|---|---|---|---|---|---|
| Qwen 3.8 FlashAlibaba | $0.16/1M | $0.47/1M | $2.54 | 991k tokens | Very fast | 84 | 76 | 78 |
| Qwen 3.8 MaxAlibaba | $2.00/1M | $6.00/1M | $32 | 1M tokens | Balanced | 93 | 85 | 88 |
| Qwen 3.7 MaxAlibaba | $2.50/1M | $7.50/1M | $40 | 1M tokens | Balanced | 89 | 83 | 89 |
Capability scores are out of 100 and reflect our own weighting of published benchmarks and production signals — see how we evaluate models. “Est. month” assumes 10M input and 2M output tokens at list price, with no batch or caching discounts applied, so treat it as a ceiling.
What each one is genuinely good at, where it falls down, and the situations we would steer you away from it — not just the headline score.
Alibaba's preview of the Qwen4 architecture — 125B parameters with only 6B active per token, at sixteen cents per million input.
You need Alibaba's maximum capability — that is Qwen 3.8 Max — or a SWE-bench Verified number.
One of the best coding-score-per-dollar picks in the catalog. Route volume work here and reserve Qwen 3.8 Max or a frontier model for the hard cases.
Released August 26, 2026. The open-weight release is Qwen3.8-Flash-Next, a preview of the Qwen4 architecture: 125B mixture-of-experts with 6B active per token, a 51B n-gram embedding table and a 4B multi-token prediction layer. Qwen 3.8 Flash is the production API version on Qwen Cloud at $0.16/$0.47.
Alibaba's largest model ever — a 2.4-trillion-parameter MoE (95B active) multimodal flagship that beat GPT-5.6 Sol on SWE-bench Pro and ranks #2 globally for vision.
You need independently verified benchmarks or Western data residency.
The strongest Chinese multimodal flagship and a legitimate SWE-bench Pro upset over GPT-5.6 Sol. If vision matters, only Fable 5-class models beat it — at 3–8x the price. Wait for independent evals before betting production on the self-reported numbers.
Announced August 3, 2026 on Alibaba Cloud Model Studio; open weights promised a week after launch. $2/$6 is first-party Model Studio pricing; cache reads from $0.17/1M. Announcement moved Alibaba stock +7% in Hong Kong.
Alibaba's first closed-weight flagship — an agent-first model with native extended thinking, built to run autonomously for up to ~35 hours firing thousands of tool calls.
Starting fresh — Qwen 3.8 Max is stronger, cheaper, and multimodal.
A capable agent-first flagship that was superseded within three months by Qwen 3.8 Max at a lower price. Choose it only if its specific extended-thinking behavior fits your agent stack; otherwise 3.8 Max is the better Qwen.
Announced at Alibaba Cloud Summit May 20, 2026. API-only on DashScope/Model Studio. Cached input $0.25/1M.
UseRightAI recommendations are based on practical decision factors people actually feel in day-to-day use.
Newsletter
Useful if you care about ranking shifts, pricing changes, or a better recommendation appearing in this decision path.
No spam. Useful updates only. Affiliate disclosures always clearly labeled.
Qwen 3.8 Flash at $0.16/1M input and $0.47/1M output tokens. SWE-bench Pro 62.5 at sixteen cents per million input.
Qwen 3.8 Flash is the best capability-per-dollar pick in Alibaba's lineup (budget score 93/100). It handles cheap high-throughput coding and reasoning well — step up to Qwen 3.7 Max only where quality visibly falls short.
Qwen 3.8 Flash costs $0.16/1M input vs $2.5/1M for Qwen 3.7 Max — a 94% saving on input tokens.
Qwen 3.8 Max — 1M tokens at $2/1M input. Context is where budget models are least compromised: you usually lose reasoning depth before you lose window size, so a cheap model is often a perfectly good choice for summarising or extracting from long documents.
No published SWE-bench Verified score, only SWE-bench Pro. An architecture preview rather than a settled flagship — Qwen 3.8 Max remains Alibaba's top-end model. Avoid it if you need Alibaba's maximum capability — that is Qwen 3.8 Max — or a SWE-bench Verified number.
On a moderate workload of 10M input and 2M output tokens, Qwen 3.8 Flash runs about $2.54 against $40.00 for Qwen 3.7 Max — a difference of $37.46 a month at the same volume. Output tokens dominate the bill on both, so the length of the responses you generate matters far more than the length of your prompts.
Mixing is almost always cheaper for the same quality. Route high-volume, low-stakes work — classification, extraction, first drafts, routine agent steps — to Qwen 3.8 Flash, and reserve Qwen 3.7 Max for the calls where a wrong answer costs real time. Teams that split this way typically cut spend substantially without a quality drop anyone notices, because most tokens in a real workload are not hard problems.