Get Started
Choosing a model
The catalog is deliberately curated: five production models spanning reasoning, coding, chat, and vision, all behind the same OpenAI-compatible endpoint.
Larger models give you more capability on complex tasks; smaller ones respond faster and cost less per token. Capability is never one-to-one across model families, so treat these recommendations as starting points and spot-check candidates on your own prompts, the playground makes side-by-side comparisons quick.
Recommendations by use case
| Use case | Start with | Also consider |
|---|---|---|
| Agentic coding & complex reasoning | GLM-5.2 | DeepSeek V4 Pro |
| Low-latency agents & high-volume pipelines | DeepSeek V4 Flash | GLM-5.3 Flash |
| General chat & assistants | GLM-5.2 | DeepSeek V4 Pro |
| Vision & image understanding | Gemma 4 31B Instruct | GLM-5.3 Flash |
| Long documents & whole-codebase context | DeepSeek V4 Pro (1M-token context) | GLM-5.2 (also 1M) |
| Low-cost, short-context tasks | Gemma 4 31B Instruct | DeepSeek V4 Flash |
Supported models
Pass the model ID as the model parameter in your requests. Context and max output are shown as rounded thousands; the exact values are at GET /models.
| Model | Context | Max output | Input | Features |
|---|---|---|---|---|
Gemma 4 31B Instructgoogle/gemma-4-31b-it | 256k | 256k | Text, Image | Tools, JSON mode |
GLM-5.2zai-org/GLM-5.2 | 1M | 1M | Text | Tools, JSON mode, Reasoning |
GLM-5.3 Flashzai-org/GLM-5.3-Flash | 1M | 1M | Text, Image | Tools, JSON mode, Reasoning |
DeepSeek V4 Prodeepseek-ai/DeepSeek-V4-Pro | 1M | 1M | Text | Tools, JSON mode, Reasoning |
DeepSeek V4 Flashdeepseek/deepseek-v4-flash-0731 | 256k | 256k | Text | Tools, JSON mode, Reasoning |
In one line each:
| Model | Best for |
|---|---|
| Gemma 4 31B Instruct | Vision and image understanding at a small-model price point |
| GLM-5.2 | General chat and long-context assistants with strong tool use and visible reasoning |
| GLM-5.3 Flash | Fast multimodal chat and high-volume agents with visible reasoning |
| DeepSeek V4 Pro | Agentic coding, hard reasoning, and whole-codebase context |
| DeepSeek V4 Flash | Fast, cost-efficient reasoning for high-volume agents and pipelines |
Pricing is per token and varies by model. Live rates for your organization, including discounted cached-input pricing, are always on the dashboard's Models page.
List models via the API
The live catalog is also served by the standard OpenAI models endpoint. Each entry's id is the value you pass as model, alongside per-model context length, modalities, and supported parameters:
curl https://inference.pearlresearch.ai/v1/models \
-H "Authorization: Bearer $PEARL_API_KEY"Migrating from closed models
Moving from a closed-model provider? Because the endpoint is OpenAI-compatible, migrating is usually a base-URL and model-name change. As a starting point:
| If you use | Start with |
|---|---|
| Frontier coding & agent models (GPT-5-class, Claude Opus-class) | DeepSeek V4 Pro |
| Mid-tier general models (GPT-4o-class, Claude Sonnet-class) | GLM-5.2 |
| Fast, low-cost models (mini/Haiku-class) | DeepSeek V4 Flash |
| Fast multimodal models (GPT-4o-mini-class) | GLM-5.3 Flash |
| Multimodal models used for image understanding | Gemma 4 31B Instruct |