# Boundless

> An inference partner that serves open models per token and tunes reserved capacity to your traffic.

Canonical: https://www.subconscious.dev/providers/boundless · By The Subconscious Team · Updated October 29, 2026

- Example models: GLM-5.3, DeepSeek V4.1 Flash, Kimi K3
- Website: https://boundless.network

## Overview

Boundless launched in September 2026 as an inference partner for AI-native startups. It sells two things. The Model API serves pre-tuned models on shared capacity, billed per token from prepaid credits with a budget per API key. Tailored inference reserves capacity for one customer, and Boundless engineers pick the model on the customer’s own evals, tune precision, batching, caching, speculative decoding and parallelism, then run the deployment against uptime, throughput and latency targets written into the contract. A free two-week evaluation tests those targets on real traffic against the customer’s current provider first.

The library lists 22 models. Open ones include DeepSeek V4 Flash at $0.04 in and $0.21 out per million tokens, GLM-5.3 Flash at $0.12 in and $0.40 out, DeepSeek V4.1 Flash at $0.20 in and $1.00 out, GLM-5.3 at $1.12 in and $3.52 out and Kimi K3 at $2.30 in and $11.40 out, most with 1M context. It also resells closed models such as GPT-6 Sol and Claude Opus 5.5 through the same API, which works with the OpenAI and Anthropic SDKs. Boundless serves several of its models on Vercel AI Gateway, keeps zero data retention, runs requests in the US or Europe, and offers abliterated models for vetted security teams through a program called Defensive Access.

## Upsides

- Engineers tune and operate a reserved deployment against targets set in the contract.
- Free two-week evaluation on your own traffic before you commit.
- One OpenAI and Anthropic compatible API for open models plus GPT and Claude.
- Zero data retention, with US or Europe processing.

## Core use cases

- Startups moving production traffic from a closed model to a tuned open one.
- Teams that want engineers to tune and run their inference for them.
- Authorized security testing on abliterated models.

## Downsides

- New company with no published benchmarks, customers or founding team.
- No fine-tuning; tailoring covers the serving stack only.
- Reserved capacity and SLAs are priced per contract.

## At a glance

| Attribute | Value |
|---|---|
| Model access | Open weights, plus closed GPT and Claude |
| Flagship models | GLM-5.3, DeepSeek V4.1 Flash, Kimi K3 |
| Speed | 342 tok/s on DeepSeek V4.1 Flash 1017 (Vercel) |
| Price | $0.04–$2.30 in, $0.21–$11.40 out per 1M (open) |
| Customization | Serving-stack tuning, no fine-tuning |
| Deployment | Serverless API, reserved tailored capacity |
| Long context | 1M context on most models |

## FAQ

### What is Boundless?

Boundless launched in September 2026 as an inference partner for AI-native startups. It sells two things. The Model API serves pre-tuned models on shared capacity, billed per token from prepaid credits with a budget per API key. Tailored inference reserves capacity for one customer, and Boundless engineers pick the model on the customer’s own evals, tune precision, batching, caching, speculative decoding and parallelism, then run the deployment against uptime, throughput and latency targets written into the contract. A free two-week evaluation tests those targets on real traffic against the customer’s current provider first.

### What is Boundless best for?

Startups moving production traffic from a closed model to a tuned open one; Teams that want engineers to tune and run their inference for them; Authorized security testing on abliterated models.

### How much does Boundless cost?

Boundless pricing at a glance: $0.04–$2.30 in, $0.21–$11.40 out per 1M (open). Rates change often, so check Boundless's pricing page before committing.

### How much context does Boundless support?

Boundless's long-context support: 1M context on most models.

### What are the downsides of Boundless?

New company with no published benchmarks, customers or founding team; No fine-tuning; tailoring covers the serving stack only; Reserved capacity and SLAs are priced per contract.

### What are the best alternatives to Boundless?

Common alternatives include Subconscious, OpenAI, Anthropic, Google Vertex AI, Amazon Bedrock. Each has a head-to-head comparison with Boundless on this site.

## Comparisons

- [Subconscious vs Boundless](https://www.subconscious.dev/compare/subconscious-vs-boundless.md)
- [OpenAI vs Boundless](https://www.subconscious.dev/compare/openai-vs-boundless.md)
- [Anthropic vs Boundless](https://www.subconscious.dev/compare/anthropic-vs-boundless.md)
- [Google Vertex AI vs Boundless](https://www.subconscious.dev/compare/google-vertex-vs-boundless.md)
- [Amazon Bedrock vs Boundless](https://www.subconscious.dev/compare/aws-bedrock-vs-boundless.md)
- [Together AI vs Boundless](https://www.subconscious.dev/compare/together-ai-vs-boundless.md)
- [Fireworks AI vs Boundless](https://www.subconscious.dev/compare/fireworks-vs-boundless.md)
- [Baseten vs Boundless](https://www.subconscious.dev/compare/baseten-vs-boundless.md)
- [Groq vs Boundless](https://www.subconscious.dev/compare/groq-vs-boundless.md)
- [Cerebras vs Boundless](https://www.subconscious.dev/compare/cerebras-vs-boundless.md)
- [DeepInfra vs Boundless](https://www.subconscious.dev/compare/deepinfra-vs-boundless.md)
- [Hugging Face Inference Providers vs Boundless](https://www.subconscious.dev/compare/hugging-face-vs-boundless.md)
- [Modal vs Boundless](https://www.subconscious.dev/compare/modal-vs-boundless.md)
- [Cloudflare Workers AI vs Boundless](https://www.subconscious.dev/compare/cloudflare-workers-ai-vs-boundless.md)
- [xAI vs Boundless](https://www.subconscious.dev/compare/xai-vs-boundless.md)
- [Mistral AI vs Boundless](https://www.subconscious.dev/compare/mistral-ai-vs-boundless.md)
- [DeepSeek vs Boundless](https://www.subconscious.dev/compare/deepseek-vs-boundless.md)
- [Moonshot AI vs Boundless](https://www.subconscious.dev/compare/moonshot-ai-vs-boundless.md)
- [Z.ai vs Boundless](https://www.subconscious.dev/compare/z-ai-vs-boundless.md)
- [Alibaba Cloud vs Boundless](https://www.subconscious.dev/compare/alibaba-cloud-vs-boundless.md)
- [Meta vs Boundless](https://www.subconscious.dev/compare/meta-vs-boundless.md)
- [Cohere vs Boundless](https://www.subconscious.dev/compare/cohere-vs-boundless.md)
- [SambaNova vs Boundless](https://www.subconscious.dev/compare/sambanova-vs-boundless.md)
- [Nebius vs Boundless](https://www.subconscious.dev/compare/nebius-vs-boundless.md)
- [Crusoe vs Boundless](https://www.subconscious.dev/compare/crusoe-vs-boundless.md)
- [fal vs Boundless](https://www.subconscious.dev/compare/fal-vs-boundless.md)
- [Novita AI vs Boundless](https://www.subconscious.dev/compare/novita-ai-vs-boundless.md)
- [Venice vs Boundless](https://www.subconscious.dev/compare/venice-vs-boundless.md)
- [Parasail vs Boundless](https://www.subconscious.dev/compare/parasail-vs-boundless.md)
- [Inference.net vs Boundless](https://www.subconscious.dev/compare/inference-net-vs-boundless.md)
- [GMI Cloud vs Boundless](https://www.subconscious.dev/compare/gmi-cloud-vs-boundless.md)
- [Thinking Machines vs Boundless](https://www.subconscious.dev/compare/thinking-machines-vs-boundless.md)
- [Sail Research vs Boundless](https://www.subconscious.dev/compare/sail-research-vs-boundless.md)
- [Morph vs Boundless](https://www.subconscious.dev/compare/morph-vs-boundless.md)
- [Relace vs Boundless](https://www.subconscious.dev/compare/relace-vs-boundless.md)
- [TypeSafe AI vs Boundless](https://www.subconscious.dev/compare/typesafe-ai-vs-boundless.md)
- [StepFun vs Boundless](https://www.subconscious.dev/compare/stepfun-vs-boundless.md)
- [Runware vs Boundless](https://www.subconscious.dev/compare/runware-vs-boundless.md)
- [StreamLake vs Boundless](https://www.subconscious.dev/compare/streamlake-vs-boundless.md)
- [Wafer vs Boundless](https://www.subconscious.dev/compare/wafer-vs-boundless.md)
- [RunInfra vs Boundless](https://www.subconscious.dev/compare/runinfra-vs-boundless.md)
- [Particle.AI vs Boundless](https://www.subconscious.dev/compare/particle-ai-vs-boundless.md)
- [Luminal vs Boundless](https://www.subconscious.dev/compare/luminal-vs-boundless.md)
- [Infron vs Boundless](https://www.subconscious.dev/compare/infron-vs-boundless.md)
- [Inception vs Boundless](https://www.subconscious.dev/compare/inception-vs-boundless.md)
- [Dolphin vs Boundless](https://www.subconscious.dev/compare/dolphin-vs-boundless.md)
- [OpenScale vs Boundless](https://www.subconscious.dev/compare/openscale-vs-boundless.md)
- [General Compute vs Boundless](https://www.subconscious.dev/compare/general-compute-vs-boundless.md)

## Sources

- [Boundless](https://boundless.network/)
- [Boundless models and pricing](https://boundless.network/models)
- [Boundless tailored inference](https://boundless.network/product/forward-deployed-inference)
- [Introducing Boundless AI](https://boundless.network/news/introducing-boundless-ai-the-inference-partner-for-ai-native-startups)
- [Boundless on Vercel AI Gateway](https://vercel.com/ai-gateway/models/providers/boundless)

Pricing and model lineups change often; figures are a snapshot.