Baseten provides on-demand frontier model APIs designed for production applications, not just experimentation. Built on the Baseten Inference Stack, these APIs deliver enterprise-grade performance and reliability with optimized inference for leading open-source models from OpenAI, DeepSeek, Meta, Moonshot AI, and Alibaba Cloud. Website: https://www.baseten.co/products/model-apis/

​
Getting an API Key

  1. Sign Up/Sign In: Go to Baseten and create an account or sign in.
  2. Navigate to API Keys: Access your dashboard and go to the API Keys section.
  3. Create a Key: Generate a new API key. Give it a descriptive name (e.g., “Cline”).
  4. Copy the Key: Copy the API key immediately and store it securely.

​
Supported Models

Cline supports all current models under Baseten Model APIs, including: For the most updated pricing, please visit: https://www.baseten.co/products/model-apis/ Reasoning Models:
  • deepseek-ai/DeepSeek-R1 - DeepSeek’s first-generation reasoning model (163K context) - $2.55/$5.95 per 1M tokens
  • deepseek-ai/DeepSeek-R1-0528 - Latest revision of DeepSeek’s reasoning model (163K context) - $2.55/$5.95 per 1M tokens
  • deepseek-ai/DeepSeek-V3.1 - Hybrid reasoning with advanced tool calling (163K context) - $0.50/$1.50 per 1M tokens
  • deepseek-ai/DeepSeek-V3-0324 - Fast general-purpose with enhanced reasoning (163K context) - $0.77/$0.77 per 1M tokens
Flagship Models:
  • openai/gpt-oss-120b (OpenAI) - 120B MoE with strong reasoning capabilities (128K context) - $0.10/$0.50 per 1M tokens
  • moonshotai/Kimi-K2-Instruct (Moonshot AI) - 1 trillion parameter model for agentic tasks (131K context) - $0.60/$2.50 per 1M tokens
  • moonshotai/Kimi-K2-Instruct-0905 (Moonshot AI) - September update with enhanced capabilities (262K context) - $0.60/$2.50 per 1M tokens
Meta Llama 4 Series:
  • meta-llama/Llama-4-Maverick-17B-128E-Instruct - High-efficiency processing (1M context!) - $0.19/$0.72 per 1M tokens
  • meta-llama/Llama-4-Scout-17B-16E-Instruct - Precise context understanding (1M context!) - $0.13/$0.50 per 1M tokens
Coding Specialists:
  • Qwen/Qwen3-Coder-480B-A35B-Instruct- Advanced coding and reasoning (262K context) - $0.38/$1.53 per 1M tokens
  • Qwen/Qwen3-235B-A22B-Instruct-2507 - Math and reasoning expert (262K context) - $0.22/$0.80 per 1M tokens

​
Configuration in Cline

  1. Open Cline Settings: Click the settings icon (⚙️) in the Cline panel.
  2. Select Provider: Choose “Baseten” from the “API Provider” dropdown.
  3. Enter API Key: Paste your Baseten API key into the “Baseten API Key” field.
  4. Select Model: Choose your desired model from the “Model” dropdown.

​
Production-First Architecture

Baseten’s Model APIs are built for production environments with several key advantages:

​
Enterprise-Grade Reliability

  • Four nines of uptime (99.99%) through active-active redundancy
  • Cloud-agnostic, multi-cluster autoscaling for consistent availability
  • SOC 2 Type II certified and HIPAA compliant for security requirements

​
Optimized Performance

  • Pre-optimized models shipped with the Baseten Inference Stack
  • Latest-generation GPUs with multi-cloud infrastructure
  • Ultra-fast inference optimized from the bottom up for production workloads

​
Cost Efficiency

  • 5-10x less expensive than closed alternatives
  • Optimized multi-cloud infrastructure for efficient resource utilization
  • Transparent pricing with no hidden costs or rate limit surprises

​
Developer Experience

  • OpenAI compatible API - migrate by swapping a single URL
  • Drop-in replacement for closed models with comprehensive observability
  • Seamless scaling from Model APIs to dedicated deployments

​
Special Features

​
Function Calling & Tool Use

All Baseten models support structured outputs, function calling, and tool use as part of the Baseten Inference Stack, making them ideal for agentic applications.

​
Reasoning Capabilities

DeepSeek models offer enhanced reasoning with step-by-step thought processes, while maintaining production-ready performance.

​
Long Context Support

  • Up to 1 million tokens for Llama 4 models (Maverick and Scout)
  • 262K tokens for Qwen3 models
  • 163K tokens for DeepSeek models
  • Perfect for code repositories and complex multi-turn conversations

​
Quantization Optimizations

Models are deployed with advanced quantization techniques (fp4, fp8, fp16) for optimal performance while maintaining quality.

​
Migration from Other Providers

Baseten’s OpenAI compatibility makes migration straightforward: From OpenAI:
  • Swap api.openai.com with inference.baseten.co/v1
  • Keep existing request/response formats
  • Benefit from significant cost savings
From Other Providers:
  • Use standard OpenAI SDK format
  • Maintain existing prompting strategies
  • Access to newer open-source models

​
Tips and Notes

  • Model Selection: Choose models based on your specific use case - reasoning models for complex tasks, coding models for development work, and flagship models for general applications.
  • Cost Optimization: Baseten offers some of the most competitive pricing in the market, especially for open-source models.
  • Context Windows: Take advantage of large context windows (up to 1M tokens) for including substantial codebases and documentation.
  • Enterprise Ready: Baseten is designed for production use with enterprise-grade security, compliance, and reliability.
  • Dynamic Model Updates: Cline automatically fetches the latest model list from Baseten, ensuring access to new models as they’re released.
  • Multi-Cloud Capacity Management (MCM): Baseten’s multi-cloud infrastructure ensures high availability and low latency globally.
  • Support: Baseten provides dedicated support for production deployments and can work with you on dedicated resources as you scale.

​
Pricing Information

Current pricing is highly competitive and transparent. For the most up-to-date pricing, visit the Baseten Model APIs page. Prices typically range from $0.10-$6.00 per million tokens, making Baseten significantly more cost-effective than many closed-model alternatives while providing access to state-of-the-art open-source models.