whichdevtool?
Menu
StacksDirectory Pricing changes Glossary
VPS and cloud servers App hosting and PaaS Managed databases Transactional email APIs Email marketing and newsletters Domain registrars CDN, DNS and edge security Monitoring, uptime and error tracking Auth and identity Frontend and full-stack frameworks Backend frameworks UI component libraries AI model APIs File and object storage Payments and billing Background jobs, queues and workflows ORMs and database toolkits

Inference provider (open-weight models)

Together AI

Prepaid inference cloud for open-weight models such as DeepSeek, Qwen, Kimi, GLM, Llama and gpt-oss, plus dedicated endpoints, fine-tuning and GPU clusters.

Visit together.ai ↗ Alternatives Paid
Together AI homepage
§1

Overview

Together AI is an inference cloud for open-weight models. Its serverless API covers families such as DeepSeek, Qwen, Kimi, GLM, Llama and OpenAI's gpt-oss, and the same account can move to dedicated endpoints, fine-tuning jobs or self-serve GPU clusters as a workload grows.

Serverless prices vary widely by model: Qwen3.8 Flash costs $0.09 in and $0.28 out per million tokens, gpt-oss-120B $0.15 / $0.60, and Kimi K3 $3 / $15. For many models the price matches or sits close to what the model's maker charges on its own API, so the main reasons to choose Together are catalogue breadth and one bill.

Billing is fully prepaid. You must buy at least $5 of credit before using the platform, there is no free trial, and when the balance reaches zero API access is suspended until you add credit. Auto-recharge can top up when the balance falls below a threshold (card payments only).

§2

Pricing and plans

Plan Price What you get
Qwen3.8 Flash $0.09 in / $0.28 out per 1M Serverless
gpt-oss-120B $0.15 in / $0.60 out per 1M Serverless
GLM-5.3-Flash $0.15 in / $0.50 out per 1M Serverless
Llama 3.3 70B $1.04 in / $1.04 out per 1M Serverless
DeepSeek V4 Pro 0813 $1.32 in / $3.96 out per 1M Serverless
Kimi K3 $3 in / $15 out per 1M Serverless
Dedicated endpoints / GPU clusters Per GPU-hour See pricing page

Per 1M tokens in USD, read 2026-09-29 on together.ai/pricing. Access requires a minimum $5 credit purchase; Together does not offer free trials. The platform is fully prepaid and API access is suspended at a zero balance; automatic credit purchases can be set for card payments. The pricing page shows Batch API price columns for some model categories, but the discount rate was not stated in the content read. One model (Ternary Bonsai 27B) was listed at $0.00.

Source: www.together.ai/pricing · prices as published, excluding tax.

§3

What it does

Serverless inference
OpenAI-compatible API for a large catalogue of open-weight models billed per token.
Dedicated endpoints
Run a model on reserved infrastructure for steadier performance and pricing.
Fine-tuning
Refine open-source models and serve the result on the same platform.
GPU clusters
Self-serve instant clusters up to large multi-GPU reservations.
Auto-recharge
Automatic credit purchase when the balance drops below a threshold you set.
§4

Strengths and limitations

Strengths

  • One account covers serverless, dedicated, fine-tuning and raw GPUs.
  • Wide choice of open-weight model families.
  • Prepaid billing caps spend at the credit you load.

Limitations

  • No free trial; a $5 minimum purchase is required.
  • API access is suspended at a zero balance unless auto-recharge is on.
  • Auto-recharge does not work when ACH is the default payment method.
  • No proprietary models from OpenAI, Anthropic or Google.
§5

Who it suits

Good for

  • Products that switch between several open-weight models
  • Teams planning to fine-tune an open model and serve it
  • Startups that may later need dedicated GPUs

Look elsewhere if

  • Hobby projects that need a free allowance
  • Apps that need proprietary frontier models
§6

Alternatives to Together AI

Tool From Free option Stages
Claude API AI model API (model maker)
Usage-based Trial
Fireworks AI Inference provider (open-weight models)
Usage-based Trial
Mistral API (Mistral Studio) AI model API (model maker)
Usage-based Free tier
Cerebras Inference Inference provider (open-weight models)
Usage-based Trial
OpenRouter Model router (unified API)
— Free tier
OpenAI API AI model API (model maker)
Usage-based Paid

All AI model APIs →

§7

Questions

Does Together AI have free credits?

No. Together's billing docs say it does not offer free trials and requires a minimum $5 credit purchase.

What happens when my Together AI balance hits zero?

API access is suspended until you add credits; auto-recharge can top up automatically below a threshold.

How much does gpt-oss-120B cost on Together AI?

$0.15 per 1M input tokens and $0.60 per 1M output tokens on 2026-09-29.

Sources read for this page