Inception
The only commercial diffusion LLMs, generating tokens in parallel for 1,000+ tok/s on standard GPUs.
- Founded
- 2024
- Example models
- Mercury 2.5, Mercury 2
By The Subconscious Team · Updated
What is Inception?
Inception builds diffusion large language models (dLLMs). Every other LLM on this list generates text one token at a time; Inception's Mercury models start from a rough draft of a whole block and refine it over a few passes, producing many tokens in parallel. That ties speed less to output length and lets the models run fast on ordinary NVIDIA GPUs rather than custom chips. The company was founded by professors from Stanford, UCLA and Cornell behind core work on diffusion, Flash Attention and DPO, and is led by Stanford's Stefano Ermon. It raised a $50M seed in November 2025 led by Menlo Ventures, with Mayfield, M12, NVentures, Snowflake and Databricks, plus Andrew Ng and Andrej Karpathy as angels.
Mercury 2.5, released in September 2026, is the flagship: a reasoning model with tunable reasoning depth, native tool use, JSON mode and a 260K context window, priced at $0.20 in and $0.75 out per million tokens. Inception reports over 1,100 tokens per second in production; Artificial Analysis measures about 640 for Mercury 2.5 and 850 for Mercury 2. Quality sits with cost-optimized models like Claude Haiku 4.5 and Gemini 3.5 Flash-Lite, not the frontier. The lineup also includes Mercury Voice, tuned for voice agents with time to first token under 170ms, and Mercury Router for routing prompts across models. The API is OpenAI-compatible and also sold through AWS Bedrock, Azure Foundry and OpenRouter, with fine-tuning and private deployments on request.
Inception pros, cons and use cases
Upsides
- Among the fastest output of any hosted LLM, without specialized silicon.
- Low prices for a reasoning model with tool use and 260K context.
- Available on AWS Bedrock and Azure Foundry for teams that buy through a cloud.
Core use cases
- Latency-critical loops: code autocomplete and edits, voice agents, real-time chat.
- High-volume summarization, extraction and classification where speed drives cost.
Downsides
- Closed weights from a single lab, so there is no choice of model family.
- Intelligence trails frontier and large open models on hard reasoning and long agent tasks.
Inception alternatives compared
Pick any row for the full head-to-head.
| Provider | Model access | Flagship models | Speed | Price | Customization | Deployment | Long context | Compare |
|---|---|---|---|---|---|---|---|---|
 Inception | Closed weights (Mercury dLLMs) | Mercury 2.5, Mercury 2 | 1,100+ tok/s (vendor); ~640 measured | $0.20 in, $0.75 out (Mercury 2.5) | Fine-tuning on request | Serverless API, Bedrock, Azure, private | 260K | |
 Subconscious | Open weights | GLM 5.3, DeepSeek V4.1 Flash | 2x faster task completion | 50–80% lower cost; billed on processed tokens | Marathon post-trained variants | Managed API, dedicated, on-prem | 5M+ effective context | Compare |
 OpenAI | Closed, plus open gpt-oss | GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna | Fast mode; Ultrafast on GPT-6 Astra | $0.10–$10 in, $0.50–$50 out per 1M | N/A | API, Azure OpenAI, Bedrock | 1.05M; 2x input past 272K | Compare |
 Anthropic | Closed | Claude Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 5.5 | Fable is the slowest tier | $0.10–$10 in, $0.50–$50 out per 1M | N/A | API, Bedrock, Vertex AI, Microsoft Foundry | 1M; Haiku 5.5 costs more past 100K | Compare |
 Google Vertex AI | Closed and open, 200+ models | Gemini 3.8 Flash, Claude, Gemma | Flash tier built for low latency | Gemini 3.8 Flash $0.75 in, $3.75 out until 2027, then $1.50/$7.50 | Custom training on GPUs or TPUs | Managed on Google Cloud | 1M on Gemini 3.8 Flash | Compare |
 Amazon Bedrock | Closed and open, 100+ models | Claude, GPT-6 Astra, Nova, DeepSeek | Latency-optimized option on some models | ~20–35% above direct; Claude and GPT at parity | Fine-tuning, Custom Model Import | Managed on AWS, AgentCore | Varies by model | Compare |
 Together AI | Open weights | Kimi K3, GLM 5.3, DeepSeek V4.1 Flash, Qwen 3.8 | 0.99s TTFT on DeepSeek V4 Pro | Parity with Fireworks and Baseten | LoRA and full SFT; RL in beta | Serverless, dedicated, GPU clusters | 512K on DeepSeek V4 Pro 0813 | Compare |
 Fireworks AI | Open weights | DeepSeek V4.1 Flash, GLM 5.3, Kimi K3 | 167–174 tok/s on DeepSeek V4 Pro | Fine-tunes served at base price | SFT, DPO, RFT; Training API | Serverless, dedicated GPUs | Full 1M on DeepSeek V4.1 Flash, GLM 5.3 | Compare |
 Baseten | Open weights, about a dozen curated | GLM 5.3, DeepSeek V4.1 Flash, Kimi K3, gpt-oss 120B | 0.49s TTFT, lowest measured | H100 about $6.50/hr dedicated | Deploy any model with Truss | Model APIs, dedicated, self-host | Varies by model | Compare |
 Groq | Open weights | GPT-OSS 120B, Qwen 3.8 27B | 500–1,000 tok/s | Near the floor on small models | No fine-tuned model hosting | GroqCloud API | Around 131K max | Compare |
 Cerebras | Open weights | GPT-OSS 120B, Qwen 3.8 27B | ~3,000 tok/s on GPT-OSS 120B | $0.35 in, $0.75 out (GPT-OSS 120B) | Unknown | Shared API, dedicated, partners | 131K on GPT-OSS 120B, 128K on Qwen 3.8 27B | Compare |
 DeepInfra | Open weights | DeepSeek V4 Flash, Llama 3.1 8B | Unknown | From $0.02 per 1M | No managed fine-tuning | Shared API, no contracts | 1M on DeepSeek V4 Pro | Compare |
 Hugging Face Inference Providers | Open weights | GLM 5.3, Kimi K3, DeepSeek V4.1 Flash | Routes to fastest provider by default | Provider rates, no markup | N/A | Serverless router; dedicated Endpoints | Up to 1M, provider-dependent | Compare |
 Modal | Bring your own weights | None hosted | ~1s container boot | Per second; H100 $3.95/hr list | Run any training code | Serverless GPU containers | Depends on the model you deploy | Compare |
 Cloudflare Workers AI | Open weights | DeepSeek V4 Pro, GLM 5.3, Kimi K3, Kimi K2.7 Code, gpt-oss 120B | Unknown | $0.011 per 1K Neurons; 10K free daily | BYO LoRA on small models (beta) | Serverless on Cloudflare network | 1M on DeepSeek V4 and Kimi K3; 262K on Kimi K2.6 and K2.7 Code | Compare |
 xAI | Closed | Grok 4.7, Grok 4.20, grok-build | ~54 tok/s on Grok 4.6 | $2 in, $6 out (Grok 4.7); 2x past 200K | Unknown | First-party API | 500K (4.7), 1M (4.20, 4.3) | Compare |
 Mistral AI | Open weights, plus closed Codestral | Mistral Large 4, Medium 3.5, Small 4, Large 3 | Unknown | $0.15–$1.50 in, $0.60–$7.50 out per 1M | Forge (enterprise); fine-tuning API deprecated | API, Azure, Bedrock, Vertex, self-host | 1M (Large 4 preview); 256K on others | Compare |
 DeepSeek | Open weights (MIT) | DeepSeek V4.1 Flash | ~209 tok/s on V4.1 Flash | Off-peak hours at half price | Open weights to fine-tune | First-party API, Hugging Face weights | 1M, 384K max output | Compare |
 Moonshot AI | Open weights, custom license | Kimi K3, Kimi K2.8 Preview | ~33 tok/s on Kimi K3 | $3 in, $15 out (Kimi K3) | Open weights to fine-tune | API, Kimi Code, OpenRouter | 1M | Compare |
 Z.ai | Open weights (GLM-5.3 License; MIT on Flash and older) | GLM-5.3, GLM-5.3-Flash | ~80 tok/s on GLM-5.3 | $1.40 in, $4.40 out (GLM-5.3); free Flash tier | Open weights; GLM-5.3 License adds review for large hosts | API, GLM Coding Plan | 1M (GLM-5.3) | Compare |
 Alibaba Cloud | Open weights incl. Max (custom license) | Qwen 3.8-Max, Qwen 3.7-Max | ~40 tok/s on Qwen 3.8-Max | $2 in, $6 out international | No fine-tuning on hosted Max | Model Studio on Alibaba Cloud | 1M (Qwen 3.8-Max) | Compare |
 Meta | Closed API; open Muse Glimmer | Muse Spark 1.3, Muse Glimmer | ~145–233 tok/s on Muse Spark 1.3 | $1.25 in, $4.25 out; Contributor tier cheaper | Open Muse Glimmer weights to fine-tune | Meta Model API (preview) | 1M | Compare |
 Cohere | Closed, plus open Command A+ | Command A+, Command A, Embed 5, Rerank 4 | 375 tok/s on Command A+ W4A4, per Cohere | $0.0375–$2.50 in, $0.15–$10 out per 1M | Enterprise fine-tuning, incl. private | API, Bedrock, Azure, OCI, VPC, on-prem | 256K on Command A; 128K on A+ | Compare |
 SambaNova | Open weights | MiniMax M3, MiniMax M2.7, GPT-OSS 120B | ~800 tok/s on MiniMax M2.7 (SN50, SemiAnalysis) | $0.22 in, $0.59 out (GPT-OSS 120B) | Unknown | SambaCloud, racks for neoclouds | Up to 1M (MiniMax M3, preview) | Compare |
 Nebius | Open weights, 60+ models | DeepSeek V4 Pro, Kimi K3, MiniMax M3, GLM | Among top hosts on throughput | From $0.06 per 1M input | Serve uploaded fine-tunes | Token Factory, dedicated, raw GPUs | Varies by model | Compare |
 Crusoe | Open weights | DeepSeek V4, GLM 5.3, Kimi K2.6, Nemotron 3 | Up to 9.9x faster TTFT vs vLLM (vendor claim) | $0.05–$1.74 in, $0.20–$4.40 out per 1M | Serverless LoRA fine-tuning | Serverless, self-serve and tailored dedicated, raw GPUs | Varies by model; cluster-wide KV cache | Compare |
 fal | Hosted media models | FLUX, Kling, Seedream | Cold starts on less popular endpoints | Per image, per video second, GPU time | LoRA training endpoints | Hosted API, serverless GPUs | Not applicable | Compare |
 Novita AI | Open weights | DeepSeek V4 Pro, Gemma 4 | ~36 tok/s on DeepSeek V4 Pro | From $0.02 per 1M; batch 50% off | Hot-swappable LoRA adapters | Serverless, GPU cloud, dedicated | Full 1M on DeepSeek V4 Pro | Compare |
 Venice | Open weights, plus proxied closed models | GLM 5.3, Kimi K3, DeepSeek V4 Pro | Unknown | $0.06–$12 in, $0.28–$60 out per 1M; DIEM staking | Unknown | Serverless API, consumer app | 1M on most current models | Compare |
 Parasail | Any Hugging Face model | Kimi K3, DeepSeek V4.1 Flash, GLM-5.3, MiniMax M3 | 600ms p99 real-time budget | Per-parameter rates; batch 50% off | Private Hugging Face repos | Serverless, elastic, dedicated, batch | Varies by model | Compare |
 Inference.net | Open, closed and custom | Schematron, ClipTagger, customer fine-tunes | Batch windows of 24h to 7 days | Free pay-as-you-go; Growth $250/mo; B200 $9.98/hr | Distill traces into custom models | Batch API, gateway, dedicated GPUs | Varies by model | Compare |
 GMI Cloud | Open and third-party models | DeepSeek V4 Flash, GLM-5.2, Google Veo | Near bare-metal performance | H100 from $2.60/GPU-hr | Unknown | Shared, autoscaling, reserved GPUs | Varies by model | Compare |
| Thinking Machines | Open weights | Inkling, Inkling-Small | Unknown | Per 1M tokens by prefill, sample, train; Inkling $1.00 in, $4.05 out | LoRA SFT and RL via Tinker | Training API, beta serverless (Inkling only) | Inkling up to 1M (256K serverless); Tinker 32K–256K | Compare |
 Sail Research | Open weights | Kimi K2.6, GLM-5.3, GPT-OSS 120B | Low-latency default; slower windows cheaper | 30–85% below other hosts by completion window | Customer LoRA fine-tunes | API plus Sailboxes | Varies by model | Compare |
 Morph | Specialist models | morph-v3-fast, morph-v3-large | 10,500+ tok/s Fast Apply | $0.80 in, $1.20 out (morph-v3-fast) | Fine-tuning offered | OpenAI-compatible API; self-hosted on Enterprise | Unknown | Compare |
 Relace | Specialist models | relace-apply-3, agentic search | ~10,000 tok/s apply | 3x+ cheaper than full rewrites | Unknown | Hosted API or self-hosted | 128K max | Compare |
 TypeSafe AI | Decision models | Jev, jev-1.13 | 70–500ms per call | $0.042 per 1M input tokens; output free | Unknown | Early-access API, also via LiteLLM | 32K | Compare |
 StepFun | Open (Apache 2.0) and API models | Step 5 Preview, Step 3.7 Flash | ~128 tok/s on Step 3.7 Flash | $0.20 in, $1.15 out (Step 3.7 Flash) | Open weights to fine-tune | First-party API, OpenRouter | Up to 1M (Step 5 Preview); 256K on Step 3.7 Flash | Compare |
 Runware | Hosted media models | Seedance 2.5, Qwen-Image-3.0 | Unknown | Images from fractions of a cent | Fine-tuned diffusion checkpoints | Unified API, serverless GPUs, bare metal | Not applicable | Compare |
 StreamLake | Proprietary coding models, plus hosted open models | KAT-Coder-Pro V2.5, KAT-Coder-Air V2.5 | Unknown | $0.74 in, $2.96 out (Pro V2.5), or KwaiKAT Coding Plan | Unknown | MaaS API, bare metal | 256K | Compare |
 Wafer | Open weights | GLM-5.3, DeepSeek V4.1 Flash | ~176 tok/s on GLM-5.3; 2–2.8x vs stock vLLM or SGLang (vendor) | Per token, prepaid credits | Agent-tuned dedicated deployments | Serverless, dedicated | Up to 1M (GLM-5.3, DeepSeek V4.1 Flash) | Compare |
 RunInfra | Open weights | Nemotron 3.5 Lightning 30B, Qwen 3.8 27B | ~541 tok/s on Nemotron 3.5 Lightning 30B (vendor); cold starts under 2s | From $0.05 in, $0.15 out; coding plans from $10 a month | Uploads up to 50 GB; auto-quantization | Model APIs, agent-built endpoints | Varies by model | Compare |
 Particle.AI | Open weights | DeepSeek V4.1 Flash, GLM 5.3 Flash | ~193 tok/s on DeepSeek V4.1 Flash | $0.10 in, $0.40 out (GLM 5.3 Flash) | Unknown | Via Vercel AI Gateway | 1M | Compare |
 Luminal | Small hosted catalog plus bring your own weights | DeepSeek V4.1 Flash, Ideogram 4.5 | 36K tok/s on GPT-OSS 120B, 8xH100 (vendor) | $0.1425 in, $0.57 out (DeepSeek V4.1 Flash) | Compiles any PyTorch or HF model | Serverless (early access), on-prem license | Unknown | Compare |
 Infron | Closed and open, 400+ models | DeepSeek, Qwen, Claude, Gemini, GPT | Unknown | Provider rates; 3–5% top-up fee | Custom deployments | Gateway API, dedicated, BYOK | Varies by model | Compare |
 Dolphin | Open weights, uncensored fine-tunes | DeepSeek V4.1 Flash, Dolphin Mistral 24B | Unknown | $0.01 in, $0.07 out (DeepSeek V4.1 Flash) | Own uncensored fine-tunes | Decentralized serverless API | Unknown | Compare |
Frequently asked questions
What is Inception?
Inception builds diffusion large language models (dLLMs). Every other LLM on this list generates text one token at a time; Inception's Mercury models start from a rough draft of a whole block and refine it over a few passes, producing many tokens in parallel. That ties speed less to output length and lets the models run fast on ordinary NVIDIA GPUs rather than custom chips. The company was founded by professors from Stanford, UCLA and Cornell behind core work on diffusion, Flash Attention and DPO, and is led by Stanford's Stefano Ermon. It raised a $50M seed in November 2025 led by Menlo Ventures, with Mayfield, M12, NVentures, Snowflake and Databricks, plus Andrew Ng and Andrej Karpathy as angels.
What is Inception best for?
Latency-critical loops: code autocomplete and edits, voice agents, real-time chat; High-volume summarization, extraction and classification where speed drives cost.
How much does Inception cost?
Inception pricing at a glance: $0.20 in, $0.75 out (Mercury 2.5). Rates change often, so check Inception's pricing page before committing.
How much context does Inception support?
Inception's long-context support: 260K.
What are the downsides of Inception?
Closed weights from a single lab, so there is no choice of model family; Intelligence trails frontier and large open models on hard reasoning and long agent tasks.
What are the best alternatives to Inception?
Common alternatives include Subconscious, OpenAI, Anthropic, Google Vertex AI, Amazon Bedrock. Each has a head-to-head comparison with Inception on this site.
Sources: Inception, Mercury 2.5 launch, Business Wire, Inception seed round, TechCrunch, Inception on Artificial Analysis. Pricing and model lineups change often; figures are a snapshot.