Bandingkan model API LLM kos rendah mengikut harga input, harga output, konteks, capability, sumber dan kesesuaian produksi.
Pemilihan API LLM murah patut bermula dengan bentuk workload, bukan sekadar kadar terendah yang dipaparkan. Untuk klasifikasi, ringkasan, routing, draf sokongan dan transformasi batch, model yang lebih murah boleh mengurangkan perbelanjaan bulanan tanpa mengubah antaramuka aplikasi. Untuk jawapan akhir, penaakulan kompleks atau ejen kod, pasukan patut membandingkan model murah dengan fallback yang lebih kuat. NextModel menghimpunkan harga, konteks, capability, sumber provider dan contoh kod di satu tempat sebelum produksi.
Asas sumber: Katalog pilihan NextModel, harga awam provider dan metadata OpenRouter apabila tersedia.
Blended price
Mulakan dengan senarai pendek, kemudian uji prompt sebenar dan bandingkan kos bulanan sebelum routing produksi.
DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...
Mistral-Small-3.2-24B-Instruct-2506 is an updated 24B parameter model from Mistral optimized for instruction following, repetition reduction, and improved function calling. Compared to the 3.1 release, version 3.2 significantly improves accuracy on...
GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs. As their most advanced small model, it is many multiples more affordable...
Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward...
Jadual perbandingan
Gunakan paparan ini untuk mengecilkan senarai produksi, membina dasar fallback atau membandingkan ekonomi model.
| Model | Provider | Input | Output | Context | Capabilities | Best for | Latency | Status | Source |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V4 Flashdeepseek/deepseek-v4-flash | DeepSeek | $0.112 / 1M tokens | $0.224 / 1M tokens | 1M | Tool callingJSON modeLong contextReasoning | low-cost Chinese tasks, long-context summary | 800-2600ms | Catalog | OpenRouter if available |
| Mistral: Mistral Small 3.2 24Bmistralai/mistral-small-3.2-24b-instruct | Mistral AI | $0.1 / 1M tokens | $0.3 / 1M tokens | 128k | Tool callingJSON modeStreamingLow cost | translation, classification | 700-2300ms | Catalog | OpenRouter if available |
| OpenAI: GPT-4o-miniopenai/gpt-4o-mini | OpenRouter | $0.15 / 1M tokens | $0.6 / 1M tokens | 128k | Tool callingVisionJSON modeLong context | low-cost chat, image understanding | 800-2400ms | Catalog | OpenRouter if available |
| Meta: Llama 4 Maverickmeta-llama/llama-4-maverick | Meta | $0.15 / 1M tokens | $0.6 / 1M tokens | 1M | JSON modeLong contextStreamingLow cost | open-model workflows, cost-sensitive long context | 950-2800ms | Catalog | OpenRouter if available |
| Google: Gemini 2.5 Flashgoogle/gemini-2.5-flash | $0.3 / 1M tokens | $2.50 / 1M tokens | 1M | Tool callingVisionJSON modeLong context | long-document summarization, image Q&A | 900-2800ms | Catalog | OpenRouter if available | |
| MoonshotAI: Kimi K2.6moonshotai/kimi-k2.6 | Moonshot AI | $0.73 / 1M tokens | $3.49 / 1M tokens | 262.1k | JSON modeLong contextStreamingTool calling | long Chinese documents, contract review | 1400-4400ms | Catalog | OpenRouter if available |
FAQ
Ia bergantung pada kadar tukaran mata wang dan panjang output. Doubao Seed 2.0 Mini kekal sebagai pilihan produksi CNY paling murah dalam katalog ini.
Tidak. Model murah sesuai untuk kerja berulang dan berisiko rendah; bagi jawapan akhir, penaakulan kompleks dan ejen kod, ia perlu dibandingkan dengan model yang lebih kuat.