Ling 3.0 Flash
inclusionai/ling-3.0-flashChatmoereasoningtoolsinclusionAI's Ling 3.0 Flash with MXFP4 expert and FP8 dense weights supports text input, reasoning, tool calling and structured JSON output. Each request has a 262,144-token context limit with at most 65,536 output tokens.
Providers
Live endpoints, cheapest first. First token is the median wait for a streamed reply; throughput covers the last day and uptime the last 24 hours.
- ImpCCHealthymxfp469% offUS
- Input
- $0.0124
- Output
- $0.0372
- Context
- 262K66K out
- First token
- 231mslast 30 min
- Throughput
- 89 tok/s
- Uptime 24h
- 100.0%
- Consensus ProtocolHealthyfp869% offUS
- Input
- $0.031
- Output
- $0.124
- Context
- 262K33K out
- First token
- 749mslast 30 min
- Throughput
- 53 tok/s
- Uptime 24h
- 99.7%
| Provider | Context | Input | Output | First token | Throughput | Uptime 24h |
|---|---|---|---|---|---|---|
| ImpCCHealthymxfp469% offUnited States | 262K66K max out | $0.0124 | $0.0372 | 231mslast 30 min | 89 tok/s | 100.0% |
| Consensus ProtocolHealthyfp869% offUnited States | 262K33K max out | $0.031 | $0.124 | 749mslast 30 min | 53 tok/s | 99.7% |
Capabilities and data
What each endpoint declares about tools, reasoning and prompt data, beside the marketplace’s own checks. The key below says what each badge means.
- ImpCCmxfp4Tools VerifiedReasoning Effort levelsData Zero retention
- Consensus Protocolfp8Tools VerifiedReasoning Partial effortData Zero retention
- Verified
- a current check called an offered tool with valid arguments
- Effort levels
- effort levels are declared, and none is failing checks
- Zero retention
- prompts are neither stored nor trained on; the claim passed marketplace review
- Partial effort
- some declared effort levels failed a check; a request naming one goes to another endpoint
Pricing
USD per million tokens.
| Provider | Input | Output | Cache read | Cache hit | Effective input |
|---|---|---|---|---|---|
| ImpCCmxfp469% off | $0.0124 | $0.0372 | $0.00248 | 91.3% | $0.003338 |
| Consensus Protocolfp869% off | $0.031 | $0.124 | $0.0155 | — | — |
Price history
Input price · USD per million tokens
Output price · USD per million tokens
Performance
Timing percentiles cover the labelled window: the last 30 minutes when it has enough samples, otherwise the last day. TTFT is the seconds to the first token, measured only on streamed requests; response is the seconds from request to last byte, measured on every request, streamed or not. Throughput is output tokens per second over the last day, counted only over streamed generations long enough to have a rate — a buffered answer has no generation phase, and a two-token one times the scheduler rather than the model. Tool calls is the share of the last day of tool requests this endpoint answered with a tool call, published once the day holds twenty of them; under that the cell is blank rather than a rate nothing measured.
| Provider | TTFT p50 | p75 | p90 | p99 | Response p50 | p75 | p90 | p99 | Throughput p50 | p75 | p90 | p99 | Tool calls 1d | Runaways 1d | Empty 1d | Capacity |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ImpCCmxfp4 | 231mslast 30 min | 237ms | 270ms | 272ms | 2.17slast 30 min | 8.57s | 11.9s | 12.4s | 89 tok/s | 115 tok/s | 126 tok/s | 141 tok/s | — | 0.0% | 0.0% | 8/8 free |
| Consensus Protocolfp8 | 749mslast 30 min | 756ms | 926ms | 957ms | 1.60slast 30 min | 4.02s | 9.94s | 14.1s | 53 tok/s | 77 tok/s | 87 tok/s | 93 tok/s | — | 0.0% | 0.0% | 16/16 free |
Time to first token by hour · last 7 days
Hourly median per provider. The line spans hours with no streamed traffic; where samples are sparse, a dot marks each hour actually timed.
Throughput by hour · last 7 days
Hourly median output tokens per second per provider, over generations long enough to have a rate. The line spans hours with no answer long enough to measure; where samples are sparse, a dot marks each hour actually measured.
Benchmarks
Scores from the benchmark board. Each figure comes from the endpoint with the most evidence for that suite, and a provisional one names how much of its suite it covers.
Uptime
Share of successful requests and probes over the last day and week. Reliability is the share of the requests a provider accepted that it answered; turned away is the share it refused with its own 429 because it was full, which uptime counts as failures.
- ImpCCmxfp41d 100.0%7d 99.4%reliability 1d 100.0%7d 99.9%turned away 1d 0.0%7d 0.5%Healthy
- Consensus Protocolfp81d 99.7%7d 97.8%reliability 1d 99.7%7d 99.7%turned away 1d 0.0%7d 1.9%Healthy
Availability by hour · last 7 days
Share of successful requests and probes per hour and provider. The line spans hours with nothing to measure; where samples are sparse, a dot marks each hour actually measured.
Activity
Prompt tokens and completion tokens per UTC day over the last 30 days. Every billed request counts, the platform's own paid probe traffic included.
curl https://api-staging.heimdallm.com/api/v1/chat/completions \ -H "Authorization: Bearer $HEIMDALL_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "inclusionai/ling-3.0-flash", "messages": [{ "role": "user", "content": "Hello" }] }'Set HEIMDALL_API_KEY to a key from your account. See the quickstart for routing suffixes and provider preferences.
About
inclusionAI's Ling 3.0 Flash with MXFP4 expert and FP8 dense weights supports text input, reasoning, tool calling and structured JSON output. Each request has a 262,144-token context limit with at most 65,536 output tokens.
- Author
- inclusionAI
- Family
- inclusionai/ling-3.0-flash
- Weights
- ImpCC mxfp4: inclusionAI/Ling-3.0-flash-fp4Consensus Protocol fp8: inclusionAI/Ling-3.0-flash-fp8
- Request parameters
- allowed_token_ids (list), cache_salt (text), chat_template_kwargs (object), frequency_penalty (-2–2), guided_choice (list), guided_grammar (text), guided_json (any value), guided_regex (text), ignore_eos (true/false), include_stop_str_in_output (true/false), logprobs (true/false), max_completion_tokens (1–65536), max_tokens (varies by provider), min_p (0–1), min_tokens (0–65536), n (1–1), parallel_tool_calls (true/false), presence_penalty (-2–2), prompt_logprobs (0–20), reasoning (object), reasoning_effort (varies by provider), repetition_penalty (varies by provider), response_format (varies by provider), return_tokens_as_token_ids (true/false; requires logprobs=true), seed (varies by provider), stop (varies by provider), stop_token_ids (list), structured_outputs (true/false), temperature (0–2), thinking (object), tool_call_dispatch (true/false), tool_choice (varies by provider), tools (varies by provider), top_k (varies by provider), top_logprobs (0–20; requires logprobs=true), top_p (varies by provider)
- Licence
- mit
- Size
- 65.6B parameters, I8 weights
- Tokenizer
- —
- Available until
- —
From the model card
What the repository states about these weights, read from inclusionAI/Ling-3.0-flash-fp4 on Nov 9, 2026.
- Languages
- —
- Papers
- —
- Datasets
- —
- Revision
- 3bae1cfAug 20, 2026
- Downloads
- 1.2K
- Likes
- 26