- Type
- Open weightsApache License 2.0
- Params
- 14.8B
- Context
- 131K
about 98K words of context
Our take
Written Sep 5, 2026Qwen3 is a 14.8-billion-parameter text model with a permissive Apache licence and a 131,072-token request limit. It suits teams who want open weights for local or hosted use without licensing restrictions, though no benchmark scores are available to verify its quality.
Pick this for local deployment or fine-tuning with full commercial freedom, or for cost-sensitive hosted inference where the cheapest tracked offer matters. Use it for long-context tasks up to 131,072 tokens if you do not need benchmark-verified quality. Skip it if you need measured quality scores, consistent pricing across providers, or guaranteed throughput.
The case for it
- Apache 2.0 licence allows commercial use, fine-tuning and redistribution without restriction.
- 131,072-token request limit is large for a model of this size.
- Lowest tracked hosted price is well under half the primary vendor's rate.
The case against it
- No benchmark scores in our data — no chat, coding, reasoning or other quality metrics held.
- Primary vendor charges several times more than the cheapest host for both input and output.
- Throughput ranges from 41 to 62 tokens per second across providers, with gaps in coverage.
How good is it?
We hold no score for this model.
So there is no figure here for everyday use, coding, agent work or writing. That is a gap in our data, not a low score.
Can you run it yourself?
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 11.7 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 19.7 GB spare means a 10% error in the size would not change the answer.
Apple M1 (8-core GPU) · 16 GB
Borderline fit on an estimated size. It leaves 0.9 GB spare on a size we calculated rather than measured, and a 10% error either way would change the answer.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked 2 hours ago — each listing carries its own date.
Cheapest of the 2 listings we can compare like for like — at 131K of context, out of 4 in the table below. One cheaper row there is outside that comparison: a different context length.
- per 1M tokens
- $0.12 in / $0.24 out
- Context served
- 131K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| NextBitint4Through OpenRouter | $0.10 / $0.22checked 2 hours ago | 41K37K max reply | 20 tok/s | No | No | Confirmed |
| OpenRouterOpenRouter's own listing | $0.12 / $0.24checked 2 hours ago | 131K | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp8Direct and through OpenRouter | $0.12 / $0.24checked 2 hours ago | 41K16K max reply through OpenRouter | 26 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| Alibaba CloudThrough OpenRouter | $0.23 / $0.91checked 2 hours ago | 131K8K max reply | 60 tok/s | No | Yesunknown period | Unknown |
DeepInfraAcross the 4 listings we hold: 3 say they do not train on prompts (1 of them only through OpenRouter), 0 say they do and 1 does not say. 2 appear in the zero-retention registry we check (1 of them only through OpenRouter); the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| NextBitint4Through OpenRouter | ✗ | ✓ | ✓ |
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| DeepInfrafp8Direct and through OpenRouter | ✓ | ✓ | ✓ |
| Alibaba CloudThrough OpenRouter | ✓ | ✓ | ✗ |
Tool calling: 3 of 4 listings say yes, 1 says no. JSON output: 4 of 4 listings say yes. Strict schema: 3 of 4 listings say yes, 1 says no.
Models people weigh against Qwen3 14B
When we formed this view
Recent changes
What moved
output −8% ($0.24 → $0.22 per 1M tokens)What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- No independent board has scored it, so we hold no quality figures at all.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 4 listings does not say whether it trains on prompts, and 1 answers only through OpenRouter, not for its own listing.
- We hold no cached-input rate for any of its listings.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
Licence and identifiers
What the licence allowsApache License 2.0, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- Qwen/Qwen3-14B
- Architecture
- Dense
- Takes in, gives back
- Text in, text out
- Catalogue slug
- qwen-qwen3-14b