Inception model product
Mercury 3
Mercury 3 is Inception’s diffusion-based large language model (dLLM), released on 2026-10-19.
Updated Oct 23, 2026. Default version: Mercury 3
LLMBoard Score59.0Mercury 3
Coverage0%10 benchmark families
Context window1MTokens
Official input price$0.10Inception API
Mercury 3 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Mercury 3 Benchmark Results
Benchmark scores for Mercury 3.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkSWE-Atlas | Score51.08% | Rank01 | Participants3 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkBrowseComp-Plus | Score55.18% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkDeepSWE 1.0 | Score46.31% | Rank02 | Participants2 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkMRCR v2 | Score75.50% | Rank02 | Participants4 | Percentile66.67% | EvidenceC | Evaluated |
| BenchmarkOmniScience | Score46.44% | Rank03 | Participants7 | Percentile66.67% | EvidenceC | Evaluated |
| BenchmarkAA-LCR | Score77.67% | Rank04 | Participants27 | Percentile88.46% | EvidenceC | Evaluated |
| BenchmarkTau3 Banking | Score30.24% | Rank05 | Participants16 | Percentile73.33% | EvidenceC | Evaluated |
| BenchmarkDeepSearchQA | Score71.33% | Rank13 | Participants13 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkGDPval-AA | Score1,237.20 points | Rank13 | Participants17 | Percentile25.00% | EvidenceC | Evaluated |
| BenchmarkMulti-Challenge | Score60.35% | Rank15 | Participants34 | Percentile57.58% | EvidenceC | Evaluated |
| BenchmarkIFBench | Score76.84% | Rank17 | Participants46 | Percentile64.44% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score89.23% | Rank35 | Participants256 | Percentile86.67% | EvidenceC | Evaluated |
Mercury 3 Arena Results
Preference and agent-evaluation results for the default version.
No Arena results
The default version does not have a matching Arena result yet.
Mercury 3 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.10 input, $0.35 output per 1M
- Official provider
- Inception
- Lowest third-party
- From $0.10 input, $0.35 output per 1M via Vercel AI Gateway
- Tracked offerings
- 6
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderVercel AI Gateway | Provider model IDinception/mercury-3 | Regionglobal | Input / 1M$0.10 | Output / 1M$0.35 | Context1M | Updated |
| ProviderOpenRouter | Provider model IDinception/mercury-3 | Regionglobal | Input / 1M$0.10 | Output / 1M$0.35 | Context1M | Updated |
| ProviderVenice AI | Provider model IDmercury-3 | Regionglobal | Input / 1M$0.10 | Output / 1M$0.35 | Context1M | Updated |
| ProviderNanoGPT | Provider model IDinception/mercury-3 | Regionglobal | Input / 1M$0.10 | Output / 1M$0.35 | Context1M | Updated |
| ProviderInception | Provider model IDmercury-3 | Regionglobal | Input / 1M$0.10 | Output / 1M$0.35 | Context1M | Updated |
| ProviderKilo Gateway | Provider model IDinception/mercury-3 | Regionglobal | Input / 1M$0.20 | Output / 1M$0.70 | Context1M | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Mercury 3 Runtime Performance
Provider-specific output speed and catalog latency for Mercury 3. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderInception | Output Speed1,068.00 tok/s | Catalog LatencyN/A | Max Input1M | Max Output131.1K | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Mercury 3 Specifications
Technical details for the model's default version.
- Version
- Mercury 3
- Released
- Oct 19, 2026
- Knowledge cutoff
- Unknown
- Parameters
- N/A
- Context window
- 1M
- Max output
- 131.1K
- Inputs
- text
- Outputs
- text
- Open weights
- No
- License
- Proprietary
Mercury 3 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionMercury 3 | Released | LLMBoard58.95 | ParametersN/A | Context1M | Max output131.1K | Open weightsNo | LicenseProprietary |
Models similar to Mercury 3
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
IN
Mercury 2
Inception
MA
Kimi K2 Thinking
Moonshot AI
BA
ERNIE 5.0
Baidu
GO
Google
Gemini 3.5 Flash Cyber
ST
Step 3.5 Flash
StepFun
IN
Ling 3.0 Flash
Inclusionai
What is Mercury 3?
Key information about Mercury 3 and its available data.
Mercury 3 is a text-in, text-out diffusion-based LLM from Inception with a 1M-token context window, up to 131,072 output tokens, tunable reasoning effort, tool calling, and structured outputs. It is available through the Inception API, OpenRouter, Vercel AI Gateway, and Baseten.
Data as of 2026-10-22.
FAQ
Common questions about Mercury 3.
When was Mercury 3 released?
Mercury 3's default version was released on Oct 19, 2026.
How much does Mercury 3 cost?
Mercury 3's official API price is $0.10 per million input tokens and $0.35 per million output tokens via Inception. The lowest tracked third-party offer starts at $0.10 input and $0.35 output via Vercel AI Gateway.
Who created Mercury 3?
Mercury 3 was created by Inception.
What is the context window for Mercury 3?
The default version has a 1M token context window.
Is Mercury 3 open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer Mercury 3?
6 provider offerings are linked to the default version.