| System | Status | Tier | Section | What it is | Pricing | Deployment | License | OTel | Latest release | GitHub | Funding | Verified |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Active | T1 | Training infrastructure | Distributed training abstraction ↗ ● | Free and open source (Apache-2.0); no paid tier ↗ 2 | Self-hosted only — pip-installable Python library (pip install accelerate); runs on local, multi-GPU, TPU and multi-node clusters ↗ 2 | Apache-2.0 ↗ ● | v1.15.0 (2026-09-09) ↗ ● | huggingface/accelerate — ~9.9k stars, last commit 2026-09 (latest release tag) ↗ ● | Hugging Face-maintained OSS library (no separate funding) ↗ ● | 2026-10-26 | ||
Active | T1 | Foundation models (substrate reference) | Open-weights frontier model family (Qwen 3 dense + MoE; 0.5B–235B) ↗ 2 | Free open-weights (Apache 2.0); Alibaba Cloud API pay-per-token ($0.10-$2 per 1M depending on tier); Together / Fireworks hosted ↗ 2 | Self-hostable (Apache 2.0 weights) + Alibaba Cloud Model Studio + Together AI + Fireworks + HuggingFace inference + Ollama ↗ 2 | Apache 2.0 (most Qwen 3 sizes); some larger MoE variants under Qwen License ↗ 2 | · | Qwen3.8 open weights: flagship Qwen3.8-2.4T-A95B + Qwen3.8-27B (Apache-2.0) (2026-08-17) ↗ 2 | github.com/QwenLM/Qwen3 ~30k+ stars; cumulative QwenLM org repos ~80k+ stars ↗ 2 | Alibaba (BABA) public; ~$200B market cap ↗ 2 | 2026-10-31 | |
Active | T1 | Foundation models (substrate reference) | Frontier foundation model family (Sonnet 4 / Opus 4.5 / Haiku 4.5) ↗ 2 | API per 1M in/out: Fable 5.1 $10/$50; Opus 5.5 $4/$20; Sonnet 5.5 $2/$10; Haiku 5.5 from $0.10/$0.50 ↗ 2 | Managed-only — Claude API/Platform plus Amazon Web Services, Google Cloud and Microsoft Azure ↗ 2 | · | · | Claude Opus 5.5 (2026-09-22); Claude Fable 5.1 (2026-09-01); Sonnet 5.5 + Haiku 5.5 current ↗ 2 | · | $65B Series H at $965B post-money (2026-05-28); S-1 filed for a proposed IPO (2026-10-26) ↗ 2 | 2026-10-31 | |
Active | T1 | Inference platforms & gateways | Ray-based AI platform ↗ 2 | Pay-as-you-go compute billed in Anyscale credits per hour (e.g. A100 ~4.96 AC/hr); $100 free credit for new users; committed contracts with volume discounts via sales ↗ 2 | Both (self-hosted + managed cloud) — Anyscale-hosted, or BYOC in your own cloud/VPC/on-prem (VMs or Kubernetes) ↗ 2 | proprietary; Ray Apache-2.0 ↗ ● | v0.26.112 (2026-10-20) — anyscale ↗ 2 | ray-project/ray — ~44.1k stars, last commit 2026-10 ↗ ● | $259M total raised through Series C 2021-12 ($99M, Addition led; $1B valuation). ↗ 2 | 2026-10-27 | ||
Active | T1 | Evaluation & observability platforms | ML + LLM observability platform ↗ 2 | AX Free $0/mo (25k spans/mo); AX Pro $50/mo (50k spans/mo); AX Enterprise custom (SaaS or self-hosted); Phoenix OSS free ↗ 2 | Both (self-hosted + managed cloud) — AX SaaS; self-hosted on AX Enterprise only ↗ 2 | proprietary; Python SDK Apache-2.0 ↗ ● | v8.61.0 (2026-10-26) — arize ↗ 2 | $130.7M total raised through Series C 2024-07 ($70M, Salesforce Ventures + Battery led). ↗ 2 | 2026-10-27 | |||
Active | T1 | Evaluation & observability platforms | OSS LLM tracing + eval ↗ 2 | Elastic-2.0 ↗ 2 | v20.26.1 (2026-10-25) ↗ 2 | Arize-ai/phoenix — ~12.0k stars, last commit 2026-10 ↗ ● | 2026-10-27 | |||||
Active | T1 | Training infrastructure | AWS ML platform ↗ 2 | Pay-as-you-go by instance type and duration, no minimum fees; free tier for new users; savings plans for commitments ↗ 2 | Managed-only — AWS ↗ 2 | proprietary; Python SDK Apache-2.0 ↗ ● | v3.27.0 (2026-10-21) — sagemaker ↗ 2 | aws/sagemaker-python-sdk — ~2.3k stars, last commit 2026-10 ↗ ● | 2026-10-27 | |||
Active | T1 | Training infrastructure | Azure ML platform ↗ 2 | No additional charge for Azure Machine Learning; pay for underlying compute and other Azure services consumed ↗ 2 | Managed-only — Azure ↗ 2 | proprietary; Python SDK (azure-ai-ml) MIT ↗ 2 | v1.35.2 (2026-10-12) — azure-ai-ml ↗ 2 | 2026-10-27 | ||||
Active | T1 | Inference platforms & gateways | ML model serving cloud ↗ 2 | Basic $0/mo pay-as-you-go; dedicated deployments billed per GPU-minute (e.g. H100 $0.10833/min); Model APIs per 1M tokens; Pro and Enterprise by quote ↗ 2 | Both (self-hosted + managed cloud) — Baseten Cloud, self-hosted in your VPC (Enterprise), or hybrid ↗ 2 | proprietary; Truss MIT ↗ ● | v0.18.36 (2026-10-21) — truss ↗ 2 | basetenlabs/truss — ~1.2k stars, last commit 2026-10 ↗ ● | $135M total raised through Series C 2024-12 ($75M, IVP led; $825M valuation). ↗ 2 | 2026-10-27 | ||
Active | T1 | Evaluation & observability platforms | Eval + observability for AI ↗ 2 | Starter $0/mo; Pro $249/mo; Enterprise custom ↗ 2 | Both (self-hosted + managed cloud) — hosted SaaS; on-prem deployment on Enterprise ↗ 2 | proprietary; SDKs open source ↗ 2 | v0.48.1 (2026-10-26) — braintrust Python SDK ↗ 2 | $36M Series A 2024-04 (a16z led; $150M valuation). ↗ 2 | 2026-10-27 | |||
Active | T1 | Foundation models (substrate reference) | Enterprise-focused frontier model family (Command A / Command R+ / Command R7B) ↗ 2 | API: Command A $2.50/$10 per 1M; Command R+ $2.50/$10; Command R7B $0.0375/$0.15; Embed + Rerank separately priced ↗ 2 | Managed cloud + customer VPC (AWS / Azure / GCP / Oracle Cloud) + on-prem; weights downloadable for non-commercial via HuggingFace ↗ 2 | Apache-2.0 (Command A+ open weights, 2026-05); earlier Command R/A weights CC-BY-NC-4.0; API proprietary ↗ 2 | · | Command A+ (command-a-plus-05-2026) (2026-05-20) ↗ 2 | github.com/cohere-ai cumulative ~5k+ stars; various model repos on HuggingFace ↗ 2 | $1.5B+ total raised; ~$5.5B valuation 2024 (Series D); Inovia, NVIDIA, Salesforce Ventures, Cisco, PSP investors ↗ 2 | 2026-10-31 | |
Acquired | T1 | Training infrastructure | Data + ML pipeline orchestrator ↗ 2 | Dagster+ Solo $10/mo, Starter $100/mo (plus per-credit and serverless compute usage), Pro custom ↗ 2 | Both (self-hosted + managed cloud) — Dagster+ Serverless or Hybrid; OSS self-hosted ↗ 2 | Apache-2.0 ↗ 2 | v1.13.27 (2026-10-26) ↗ 2 | dagster-io/dagster — ~16.3k stars, last commit 2026-10 ↗ ● | Acquired by Prefect (announced 2026-07-13; terms not disclosed); previously $47M total raised through Series B 2022-05 ($33M, Sapphire Ventures led). ↗ 2 | 2026-10-29 | ||
Active | T1 | Foundation models (substrate reference) | Open-weights reasoning-first frontier model family (R1 / V3 / V3.1) ↗ 2 | API per 1M in (cache miss)/out, peak: V4-Pro $1.32/$3.96, V4.1-Flash $0.30/$1.20; off-peak 50% lower; weights free ↗ 2 | Self-hostable (MIT weights) + hosted DeepSeek API + Together / Fireworks / HuggingFace inference ↗ 2 | MIT (R1 + V3 weights) ↗ 2 | · | DeepSeek-V4.1-Flash-1017 (2026-10-17); DeepSeek-V4-Pro GA (2026-08-13) ↗ 2 | github.com/deepseek-ai/DeepSeek-V3 ~94k stars; github.com/deepseek-ai/DeepSeek-R1 ~88k stars ↗ 2 | Self-funded via High-Flyer (parent quant fund); no external VC disclosed; estimated <$100M training cost (vs $1B+ for peers) ↗ 2 | 2026-10-31 | |
Active | T1 | Training infrastructure | ZeRO-based distributed training ↗ ● | Self-hosted only — library ↗ 2 | Apache-2.0 ↗ ● | v0.19.8 (2026-10-20) ↗ 2 | deepspeedai/DeepSpeed — ~43.3k stars, last commit 2026-10 ↗ ● | 2026-10-28 | ||||
Active | T1 | Training infrastructure | Data version control ↗ 2 | Self-hosted only — CLI and VS Code extension ↗ ● | Apache-2.0 ↗ 2 | v3.67.1 (2026-03-31) ↗ 2 | treeverse/dvc — ~15.9k stars, last commit 2026-10 ↗ ● | Iterative.ai raised $20M Series A 2022 (True Ventures led); also runs CML, MLEM, Studio. ↗ 2 | 2026-10-29 | |||
Active | T1 | Robotics foundation models & agent stacks | Humanoid robot maker (Figure 01 / Figure 02 / Helix VLA model) ↗ 2 | · | Direct customer deployment (humanoid robot) ↗ 2 | · | · | Helix 2.5 (2026-09-17) ↗ 2 | · | $675M Series B Feb-2024 ($2.6B val); Series C over $1B committed at $39B post-money, Sep-2025 (Parkway Venture Capital led) ↗ 2 | 2026-10-29 | |
Active | T1 | Inference platforms & gateways | Fast OSS LLM inference cloud ↗ 2 | Serverless pay-per-token; on-demand GPUs $8/hr (H100/H200), $13/hr (B200), $15/hr (B300); fine-tuning from $0.50 per 1M training tokens (LoRA SFT up to 16B) ↗ 2 | proprietary; Python SDK Apache-2.0 ↗ ● | v1.2.30 (2026-10-21) — fireworks-ai ↗ 2 | $77M total raised through Series B 2024-07 ($52M, Sequoia led; $552M valuation). ↗ 2 | 2026-10-27 | ||||
Active | T1 | Training infrastructure | Fully Sharded Data Parallel ↗ 2 | Self-hosted only — library (part of PyTorch) ↗ 2 | BSD-3-Clause (PyTorch) ↗ 2 | v2.14.1 (2026-09-30) — torch ↗ 2 | pytorch/pytorch — ~105k stars, last commit 2026-10 ↗ ● | 2026-10-27 | ||||
Acquired | T1 | Evaluation & observability platforms | Eval + observability + guardrails ↗ 2 | Splunk Agent Observability from $100/mo per 1.2M spans billed annually (includes 500M Luna tokens); free edition available ↗ 2 | Both (self-hosted + managed cloud) — SaaS, VPC or on-premises ↗ 2 | proprietary; Python SDK Apache-2.0 ↗ ● | v2.7.0 (2026-10-14) — galileo ↗ 2 | rungalileo/galileo-python — ~22 stars, last commit 2026-10 ↗ ● | $45M Series B Oct-2024 (Scale Venture Partners led); acquired by Cisco (announced 2026-04-09, completed by 2026-05-22) ↗ 2 | 2026-10-28 | ||
Active | T1 | Robotics foundation models & agent stacks | Robotics foundation model from Google DeepMind ↗ 2 | · | Trusted-tester via partners; Vertex AI / Cloud planned ↗ 2 | · | · | Gemini Robotics 2 (with Gemini Robotics ER 2 and On-Device 2; date not stated on the model page) ↗ 2 | · | Google DeepMind (Alphabet parent) ↗ 2 | 2026-10-29 | |
Active | T1 | Foundation models (substrate reference) | Frontier foundation model family (Gemini 3 Pro / Flash / Nano) ↗ 2 | Paid per 1M in/out: Gemini 4 Argon $2/$10 intro to 2026-11-30, then $4/$20 (no free tier); Gemini 3.8 Flash $0.75/$3.75 (free tier available) ↗ 2 | Managed cloud (Google AI Studio + GCP Vertex AI); on-prem via Gemini-Distributed-Cloud + Gemma open-weights ↗ 2 | · | · | Gemini 4 Argon (GA 2026-10-13); Gemini 3.8 Flash (2026-09-02) ↗ 2 | · | Alphabet (GOOGL) public; ~$2T+ market cap; Google AI capex $75B+ 2025 ↗ 2 | 2026-10-31 | |
Renamed | T1 | Training infrastructure | GCP ML training ↗ 2 | Managed-only — Google Cloud service ↗ 2 | proprietary; Python SDK Apache-2.0 ↗ ● | v2.6.0 (2026-10-20) — google-cloud-aiplatform ↗ 2 | googleapis/python-aiplatform — ~910 stars, last commit 2026-10 ↗ ● | 2026-10-29 | ||||
Active | T1 | Training infrastructure | Dataset library + Hub ↗ 2 | Self-hosted only — library ↗ 2 | Apache-2.0 ↗ 2 | v5.1.0 (2026-10-05) ↗ 2 | huggingface/datasets — ~22.1k stars, last commit 2026-10 ↗ ● | 2026-10-27 | ||||
Active | T1 | Robotics foundation models & agent stacks | OSS robotics platform on HuggingFace (datasets + models + policies) ↗ 2 | OSS free; HF Hub free + paid tiers ↗ 2 | Self-hosted only — library (pip install lerobot) ↗ 2 | Apache-2.0 ↗ ● | · | v0.6.1 (2026-08-03) ↗ 2 | huggingface/lerobot — ~28.5k stars, last commit 2026-10 ↗ ● | Hugging Face parent — $235M Series D 2023 ($4.5B val) ↗ 2 | 2026-10-27 | |
Active | T1 | Foundation models (substrate reference) | SSM-Transformer hybrid + 256K KV-bypass context ↗ 2 | Free trial ($10 credit, no CC required); pay-as-you-go per token; enterprise custom pricing with VPC/on-prem ↗ 2 | Both (AI21 Studio cloud; Azure/AWS Bedrock/GCP Vertex AI; private VPC; on-premises) ↗ 2 | Apache-2.0 (Jamba2); earlier Jamba 1.5–1.7 under Jamba Open Model License ↗ 2 | Jamba2 3B / Jamba2 Mini (52B, 12B active) (2026-01-08) ↗ 2 | not applicable — open weights on Hugging Face (huggingface.co/ai21labs) ↗ 2 | $608M total Series D · 2025-05 ↗ 2 | 2026-10-22 | ||
Active | T1 | Training infrastructure | Data labelling platform ↗ 2 | Community free; Starter Cloud $99/mo for up to 12 users (+$49/user); Enterprise custom ↗ 2 | Both (self-hosted + managed cloud) — Community self-host; Starter Cloud hosted; Enterprise cloud or on-prem ↗ 2 | Apache-2.0 (Community); proprietary Enterprise ↗ 2 | v1.23.2 (2026-09-29) ↗ 2 | HumanSignal/label-studio — ~28.6k stars, last commit 2026-10 ↗ ● | HumanSignal raised $25M Series A in 2022-06 (Redpoint Ventures led); Label Studio OSS is Apache-2.0. ↗ 2 | 2026-10-27 | ||
Active | T1 | Training infrastructure | Data factory for AI ↗ 2 | proprietary; Python SDK Apache-2.0 ↗ ● | v7.17.0 (2026-10-22) — labelbox SDK ↗ 2 | Labelbox/labelbox-python — ~158 stars, last commit 2026-10 ↗ ● | $100M+ Series E 2026-10-14 (equity plus credit facility; $1.25B pre-money; Infinitum Partners led, with Wipro Ventures, a16z, B Capital); previously $188.9M total raised (Series D $110M, 2022-01; led by SoftBank Vision Fund 2). ↗ 2 | 2026-10-29 | ||||
Active | T1 | Inference platforms & gateways | Universal LLM API + Proxy ↗ ● | OSS free; Enterprise custom pricing (30-day self-hosted trial key) ↗ 2 | Self-hosted only — SDK and proxy; Enterprise self-hosted or air-gapped ↗ 2 | MIT (except enterprise/ directory, commercial license) ↗ 2 | v1.106.4 (2026-10-25) ↗ 2 | BerriAI/litellm — ~62.1k stars, last commit 2026-10 ↗ ● | BerriAI raised $1.6M seed 2023 (Y Combinator W23); commercial steward of LiteLLM OSS. ↗ 2 | 2026-10-27 | ||
Active | T1 | Training infrastructure | Tensor/pipeline parallel LLM training ↗ ● | Self-hosted only — reference training scripts plus library ↗ ● | BSD-3-Clause ↗ ● | v0.19.2 (2026-09-18) — megatron-core ↗ 2 | NVIDIA/Megatron-LM — ~18.3k stars, last commit 2026-10 ↗ ● | 2026-10-27 | ||||
Active | T1 | Foundation models (substrate reference) | Open-weights frontier model family (Llama 4 Scout / Maverick / Behemoth) ↗ 2 | Free for self-hosting (Llama community license, <700M MAU); hosted via Together $0.20-$3/M (Maverick); AWS Bedrock + Azure AI consumption ↗ 2 | Both (self-hosted + managed cloud) — open weights (llama.com / Hugging Face); hosted via AWS Bedrock, Azure AI, Together, Groq, Fireworks, Databricks ↗ 2 | Llama 4 community license (Meta-custom; commercial OK <700M MAU) ↗ 2 | · | No Llama release found after Llama 4 Scout + Maverick (2025-04-05); successor Muse family: Muse Spark (2026-04-08, closed), Muse Spark Open weights (2026-10-15) ↗ 2 | github.com/meta-llama/llama (combined Llama repos); ~58k stars ↗ 2 | Meta (META) public; ~$1.5T market cap; AI infra capex $60-65B 2025 ↗ 2 | 2026-10-31 | |
Active | T1 | Foundation models (substrate reference) | Frontier model family — Mistral Large 2 + Mixtral 8x22B + Mistral Small 3 ↗ 2 | API: Mistral Medium 3.5 $1.50 in / $7.50 out per 1M; Mistral Large 3 $0.50 / $1.50; batch 50% off; open weights free ↗ 2 | Both (self-hosted + managed cloud) — La Plateforme / Mistral Studio API, open weights for self-hosting, plus Azure / AWS / GCP marketplaces ↗ 2 | Mixed: Mistral Research License MRL-0.1, non-OSI (Mistral Large 2 / 2407 and 2411 weights; commercial use needs a separate licence); Mistral Non-Production License (Codestral 22B); Modified MIT (Medium 3.5 weights); Apache-2.0 (Mixtral 8x22B, Small / Large 3 / other open models); some models API-only proprietary ↗ 2 | · | Mistral Medium 3.5 (2026-04-29); Mistral Large 4 in public preview (2026-10), weights not yet released ↗ 2 | github.com/mistralai/mistral-inference + various model repos; cumulative ~30k+ stars ↗ 2 | €2.4B+ total raised (~$2.7B USD); €11.7B valuation Series C Sept-2025 (ASML lead); a16z, General Catalyst, DST, Lightspeed, Nvidia investors ↗ 2 | 2026-10-31 | |
Active | T1 | Training infrastructure | Open ML lifecycle platform ↗ 2 | Apache-2.0 ↗ 2 | v3.17.2 (2026-10-21) — mlflow ↗ 2 | mlflow/mlflow — ~28.6k stars, last commit 2026-10 ↗ ● | OSS project — sponsored by Databricks (Databricks raised $43.2B at $62B valuation 2024-12, $10B Series J 2024-12). ↗ 2 | 2026-10-27 |