Embed Models by Cohere
8 models2023–2025Up to 128K ctxFrom $0.1/1M input
About
Embed is a family of 8 AI models by Cohere, released between 2023 and 2025.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
8 in view
Embed v4.0Current
Use when the workload needs embedding, 128K context, and multimodal inputs.
2025-04embedding128K contextmultimodal inputs
Cohere Embed EnglishCurrent
Use when provider availability and model metadata match the workload.
2024-01
Cohere Embed MultilingualCurrent
Use when provider availability and model metadata match the workload.
2024-01
Embed English v3.0Current
Use when the workload needs embedding, 512 context, and multimodal inputs.
2023-11embedding512 contextmultimodal inputs
Embed English Light v3.0Current
Use when the workload needs embedding, 512 context, and multimodal inputs.
2023-11embedding512 contextmultimodal inputs
Embed Multilingual v3.0Current
Use when the workload needs embedding, 512 context, and multimodal inputs.
2023-11embedding512 contextmultimodal inputs
Use when the workload needs embedding, 512 context, and multimodal inputs.
2023-11embedding512 contextmultimodal inputs
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Embed v4.0 | Use when the workload needs embedding, 128K context, and multimodal inputs. | 2025-04 | embedding128K contextmultimodal inputs | Current |
| Cohere Embed v4 | Use when provider availability and model metadata match the workload. | 2024-06 | — | Current |
| Cohere Embed English | Use when provider availability and model metadata match the workload. | 2024-01 | — | Current |
| Cohere Embed Multilingual | Use when provider availability and model metadata match the workload. | 2024-01 | — | Current |
| Embed English v3.0 | Use when the workload needs embedding, 512 context, and multimodal inputs. | 2023-11 | embedding512 contextmultimodal inputs | Current |
| Embed English Light v3.0 | Use when the workload needs embedding, 512 context, and multimodal inputs. | 2023-11 | embedding512 contextmultimodal inputs | Current |
| Embed Multilingual v3.0 | Use when the workload needs embedding, 512 context, and multimodal inputs. | 2023-11 | embedding512 contextmultimodal inputs | Current |
| Embed Multilingual Light v3.0 | Use when the workload needs embedding, 512 context, and multimodal inputs. | 2023-11 | embedding512 contextmultimodal inputs | Current |
Release Timeline
4 release groups2025-04
1 current
Embed v4.0
Currentembedding128K contextmultimodal inputs
2024-06
1 current
Current
2024-01
2 current
Current
Current
2023-11
4 current
Embed English Light v3.0
Currentembedding512 contextmultimodal inputs
Embed English v3.0
Currentembedding512 contextmultimodal inputs
Embed Multilingual Light v3.0
Currentembedding512 contextmultimodal inputs
Embed Multilingual v3.0
Currentembedding512 contextmultimodal inputs
Specifications(8 models)
| Model | Released | Context | Multimodal |
|---|---|---|---|
| Embed v4.0 | 2025-04 | 128k | Yes |
| Cohere Embed v4 | 2024-06 | — | No |
| Cohere Embed English | 2024-01 | — | No |
| Cohere Embed Multilingual | 2024-01 | — | No |
| Embed English v3.0 | 2023-11 | 512 | Yes |
| Embed English Light v3.0 | 2023-11 | 512 | Yes |
| Embed Multilingual v3.0 | 2023-11 | 512 | Yes |
| Embed Multilingual Light v3.0 | 2023-11 | 512 | Yes |
Available From(2 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Embed English v3.0 | Microsoft Foundry | $0.1 | — | Serverless |
| Embed Multilingual v3.0 | Microsoft Foundry | $0.1 | — | Serverless |
| Embed v4.0 | Microsoft Foundry | $0.12 | — | Serverless |
Frequently Asked Questions
- What is Embed used for?
- Embed is used for embedding and vision and multimodal work. The family description and listed model capabilities point to those workloads as the best fit.
- How does Embed compare to Command?
- Embed by Cohere is strongest where you need embedding, while Command by Cohere is the closest related family to check for multilingual. Embed has 8 listed variants and reaches up to 128K context, while Command reaches up to 256K context, so compare the specs and pricing tables before choosing a production model.
- Which Embed model should I use?
- For the lowest listed input price, start with Embed English v3.0 through Microsoft Foundry at $0.1/1M input tokens. For the most capable/latest local choice, evaluate Embed v4.0 with 128K context and multimodal inputs.
Models(8)
Embed v4.0
2025-04128k1 provider
Multimodal
Cohere Embed v4
2024-061 provider
Cohere Embed English
2024-011 provider
Cohere Embed Multilingual
2024-011 provider
Embed English v3.0
2023-115121 provider
Multimodal
Embed English Light v3.0
2023-11512
Multimodal
Embed Multilingual v3.0
2023-115121 provider
Multimodal
Embed Multilingual Light v3.0
2023-11512
Multimodal