StableLM 2 Models by Stability AI
Details
Capabilities
About
The StableLM 2 family, developed by Stability AI, features a series of open-source large language models (LLMs) designed for both strong performance and efficiency, making them compatible with a variety of hardware. This family includes models with different parameter sizes, such as the 1.6 billion parameter StableLM 2 1.6B and the 12 billion parameter StableLM 2 12B, providing flexibility for diverse computational needs. They are trained on multilingual data from languages like English, Spanish, German, Italian, French, Portuguese, and Dutch. These models are available in both base and instruction-tuned versions, catering to a wide range of applications. Released under licenses allowing commercial and non-commercial use, they come with certain terms and conditions 413.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs 4k context, 12B parameters, and structured outputs.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Stable LM 2 Zephyr 12B | Use when the workload needs 12B parameters. | 2024-07 | 12B parameters | Current |
| DeepInfra Stable LM 2 12B | Use when the workload needs 4k context, 12B parameters, and structured outputs. | 2024-06 | 4k context12B parametersstructured outputs | Current |
| Stable LM 2 12B | Use when the workload needs 12B parameters. | 2024-04 | 12B parameters | Current |
| Stable LM 2 1.6B | Use when the workload needs 1.6B parameters. | 2024-04 | 1.6B parameters | Current |
| Stable LM 2 Zephyr 1.6B | Use when the workload needs 1.6B parameters. | 2024-04 | 1.6B parameters | Current |
Release Timeline
3 release groupsSpecifications(5 models)
| Model | Released | Context | Parameters | Structured Outputs |
|---|---|---|---|---|
| Stable LM 2 Zephyr 12B | 2024-07 | — | 12B | No |
| DeepInfra Stable LM 2 12B | 2024-06 | 4k | 12B | Yes |
| Stable LM 2 12B | 2024-04 | — | 12B | No |
| Stable LM 2 1.6B | 2024-04 | — | 1.6B | No |
| Stable LM 2 Zephyr 1.6B | 2024-04 | — | 1.6B | No |
Available From(3 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Stable LM 2 Zephyr 1.6B | Fireworks AI | $0.1 | $0.1 | Provisioned |
| DeepInfra Stable LM 2 12B | DeepInfra | $0.1 | $0.3 | Serverless |
Frequently Asked Questions
- What is StableLM 2 used for?
- StableLM 2 is used for structured outputs and coding. The family description and listed model capabilities point to those workloads as the best fit.
- How does StableLM 2 compare to Stable Virtual Assistant?
- StableLM 2 by Stability AI is strongest where you need structured outputs, while Stable Virtual Assistant by Stability AI is the closest related family to check for adjacent model selection. StableLM 2 has 5 listed variants and reaches up to 4k context, so compare the specs and pricing tables before choosing a production model.
- Which StableLM 2 model should I use?
- For the lowest listed input price, start with DeepInfra Stable LM 2 12B through DeepInfra at $0.1/1M input tokens. For the most capable/latest local choice, evaluate DeepInfra Stable LM 2 12B with 4k context and structured outputs.
