Ling
inclusionAI's open mixture-of-experts models, priced near zero.
Ling is inclusionAI's open-weight family: mixture-of-experts models that activate a small slice of their parameters per token, which is how the Flash models land at a price per million tokens most people would call a rounding error. There are vision and finance-tuned variants alongside the general one, and a 262K context. A sensible pick for high-volume work — classifying, tagging, summarising — where a frontier model is overkill.
At a glance — Ling 3.0 Flash VL
- Cost
- $0.021 in · $0.062 out per 1M tokensBudget
- Memory
- 262K tokens — about 400 pages
- Reads
- Text · Images · Video
- Can do
- Use tools (agents)
- Structured output
- Thinking mode
- Good at
- Coding: better than 66% · Agentic work: better than 65% · Overall: better than 46%
- Reliability
- 99.92% uptime, last 24h
- Released
- Sep 10, 2026
Advanced details
- API model ID
inclusionai/ling-3.0-flash-vl- Max output
- 33K tokens
- Cached input
- $0.004 per 1M
- Tokenizer
- Other
- Coding index
- 57.0
- Agentic work index
- 28.7
- Overall index
- 24.6
Providers
The same model, hosted by different companies. OpenRouter picks one per request and falls back to the next if it fails.
| Provider | Input /1M | Output /1M | Cached /1M | Max output | Uptime 24h |
|---|---|---|---|---|---|
| Novita · bf16 | $0.021 | $0.062 | $0.004 | 33K | 99.92% |
| DeepInfra · fp16 | $0.06 | $0.18 | $0.012 | 33K | 99.61% |
Supported parameters
- frequency_penalty
- include_reasoning
- logit_bias
- logprobs
- max_tokens
- min_p
- presence_penalty
- reasoning
- repetition_penalty
- response_format
- seed
- stop
- structured_outputs
- temperature
- tool_choice
- tools
- top_k
- top_logprobs
- top_p
Live from OpenRouter, refreshed hourly. Scores from Artificial Analysis.
Key info
- Pricing
- Open source
- Category
- Models
- Best for
- Expert
- Made by
- inclusionAI
- Weights
- Open
- Good for
- High-volume, low-value calls