Batch variants (cheaper, slower)
If no user is staring at a spinner, you are paying for speed you do not need. A batch variant runs the same model behind the same endpoint at a lower rate, with latency measured in minutes rather than seconds. Good for overnight jobs, dataset labelling and bulk content; wrong for anything interactive.
42 models in this category · prices in toman per 1M tokens
| Model | Capabilities | Context | Input / 1M | Output / 1M | ≈ 1,000 words |
|---|---|---|---|---|---|
Claude Fable 5.1Batchanthropic/claude-fable-5.1:batch | Image inputTool callingReasoningJSON output | 1,000,000 | 1,731,097 | 8,655,484 | 13,503 |
Claude Haiku 5.5Batchanthropic/claude-haiku-5.5:batch | Image inputTool callingReasoningJSON output | 1,000,000 | 17,311 | 86,555 | 135 |
Claude Opus 5.5Batchanthropic/claude-opus-5.5:batch | Image inputTool callingReasoningJSON output | 1,000,000 | 692,439 | 3,462,194 | 5,401 |
Claude Opus 5Batchanthropic/claude-opus-5:batch | Image inputTool callingReasoningJSON output | 1,000,000 | 865,548 | 4,327,742 | 6,751 |
Claude Sonnet 5.5Batchanthropic/claude-sonnet-5.5:batch | Image inputTool callingReasoningJSON output | 1,000,000 | 346,219 | 1,731,097 | 2,701 |
 DeepSeek V4.1 FlashBatchdeepseek/deepseek-v4.1-flash:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 38,777 | 116,330 | 202 |
 Gemini 3 Flash PreviewBatchgoogle/gemini-3-flash-preview:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 86,555 | 519,329 | 788 |
Gemini 3.1 Flash LiteBatchgoogle/gemini-3.1-flash-lite:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 43,277 | 259,665 | 394 |
Gemini 3.1 Pro PreviewBatchgoogle/gemini-3.1-pro-preview:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 346,219 | 2,077,316 | 3,151 |
Gemini 3.5 Flash LiteBatchgoogle/gemini-3.5-flash-lite:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 51,933 | 432,774 | 630 |
Gemini 3.5 FlashBatchgoogle/gemini-3.5-flash:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 259,665 | 1,557,987 | 2,363 |
Gemini 3.6 FlashBatchgoogle/gemini-3.6-flash:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 129,832 | 649,161 | 1,013 |
Gemini 3.7 FlashBatchgoogle/gemini-3.7-flash:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 129,832 | 649,161 | 1,013 |
Gemini 3.8 FlashBatchgoogle/gemini-3.8-flash:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 129,832 | 649,161 | 1,013 |
Gemini 4 ArgonBatchgoogle/gemini-4-argon:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 346,219 | 1,731,097 | 2,701 |
 Mistral Medium 3.5Batchmistralai/mistral-medium-3-5:batch | Image inputTool callingReasoningJSON output | 262,144 | 259,665 | 1,298,323 | 2,025 |
Kimi K3Batchmoonshotai/kimi-k3:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 789,380 | 3,946,901 | 6,157 |
GPT-5 NanoBatchopenai/gpt-5-nano:batch | Image inputTool callingReasoningJSON output | 400,000 | 8,655 | 69,244 | 101 |
GPT-5.6 Luna ProBatchopenai/gpt-5.6-luna-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 34,622 | 207,732 | 315 |
GPT-5.6 LunaBatchopenai/gpt-5.6-luna:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 34,622 | 207,732 | 315 |
GPT-5.6 Sol ProBatchopenai/gpt-5.6-sol-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 692,439 | 3,462,194 | 5,401 |
GPT-5.6 SolBatchopenai/gpt-5.6-sol:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 692,439 | 3,462,194 | 5,401 |
GPT-5.6 Terra ProBatchopenai/gpt-5.6-terra-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 346,219 | 2,077,316 | 3,151 |
GPT-5.6 TerraBatchopenai/gpt-5.6-terra:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 346,219 | 2,077,316 | 3,151 |
GPT-6 Astra ProBatchopenai/gpt-6-astra-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 1,731,097 | 8,655,484 | 13,503 |
GPT-6 AstraBatchopenai/gpt-6-astra:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 1,731,097 | 8,655,484 | 13,503 |
GPT-6 Luna ProBatchopenai/gpt-6-luna-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 17,311 | 86,555 | 135 |
GPT-6 LunaBatchopenai/gpt-6-luna:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 17,311 | 86,555 | 135 |
GPT-6 Sol ProBatchopenai/gpt-6-sol-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 346,219 | 1,731,097 | 2,701 |
GPT-6 SolBatchopenai/gpt-6-sol:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 346,219 | 1,731,097 | 2,701 |
GPT-6.1 Sol ProBatchopenai/gpt-6.1-sol-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 346,219 | 1,731,097 | 2,701 |
GPT-6.1 SolBatchopenai/gpt-6.1-sol:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 346,219 | 1,731,097 | 2,701 |
GPT-6.1 Terra ProBatchopenai/gpt-6.1-terra-pro:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 173,110 | 865,548 | 1,350 |
GPT-6.1 TerraBatchopenai/gpt-6.1-terra:batch | Image inputTool callingReasoningJSON output | 1,050,000 | 173,110 | 865,548 | 1,350 |
gpt-oss-120bBatchopenai/gpt-oss-120b:batch | Tool callingReasoningJSON output | 131,072 | 10,248 | 47,086 | 75 |
gpt-oss-20bBatchopenai/gpt-oss-20b:batch | Tool callingReasoningJSON output | 131,072 | 8,309 | 38,777 | 61 |
 Qwen3.7 PlusBatchqwen/qwen3.7-plus:batch | Image inputTool callingReasoningJSON output | 262,144 | 79,630 | 318,522 | 518 |
Grok 4.3Batchx-ai/grok-4.3:batch | Image inputTool callingReasoningJSON output | 1,000,000 | 346,219 | 692,439 | 1,350 |
Grok 4.6Batchx-ai/grok-4.6:batch | Image inputTool callingReasoningJSON output | 500,000 | 553,951 | 1,661,853 | 2,881 |
Grok 4.7Batchx-ai/grok-4.7:batch | Image inputTool callingReasoningJSON output | 500,000 | 553,951 | 1,661,853 | 2,881 |
GLM 5.3 FlashBatchz-ai/glm-5.3-flash:batch | Image inputTool callingReasoningJSON output | 1,048,576 | 17,311 | 62,319 | 104 |
GLM 5.3Batchz-ai/glm-5.3:batch | Tool callingReasoningJSON output | 1,048,576 | 155,799 | 692,439 | 1,103 |
Other categories
Models that hold up in a real codebaseCheap models for high-volume workFree models, shared capacityModels that read imagesReasoning models, and what they costLong-context models for documents and codebasesModels that call your tools reliablyModels that return imagesEmbedding models for semantic search