Size Your LLM Hardware & API Bill
Exact KV & LoRA memory sizing from HuggingFace config.json, real-workload API billing, and 12-month GPU break-even curves.
Price & Model Release Radar
Sol priced at $2/$10 per 1M tokens for daily coding; Astra flagship at $8/$40
Off-peak window cuts V4.1-Flash to $0.15/M input ($0.015/M cached) and $0.60/M output
Fits 24GB GPUs cleanly at Q4_K_M; Bailian Omni-Flash API at ¥0.76 / ¥2.60 per 1M
1M multimodal context window with $1.50/M input and $7.50/M output
Zero-cost developer API tier with CN direct access and tool calling support
Repeated repository context billed at 10% of base input rate under Prompt Caching
Overall Top 10
All leaderboards →Latest Releases
All 270 →- 11/18
Gemini 4 Neon $3/$15
- 11/17
GPT-6.2 Sol $2/$10
- 11/17
GPT-6.2 Astra $8/$40
- 11/16
Hunyuan Hy4 Flash (混元 4 极速) $0.035/$0.141
- 11/13
Ling 3.2 Flash Free
- 11/12
Gemini 4 Hydrogen $1.5/$7.5
- 11/10
ERNIE-Light-DT-1.0 $1.13/$3.94
- 11/10
ERNIE-5.5-Flash (280B MoE) Free
- 11/10
Gemini 4 Carbon $2/$10
- 11/09
Muse Ember $0.1/$0.5
Decision Tools
VRAM & LoRA Calculator →
Inference fit, Batch 1–32 concurrency & QLoRA/LoRA sizing
API Pricing & Log Replay →
Compare 309 rates or drop CSV/JSONL logs to replay real bills
Advisor & Break-Even →
Budget hardware builds + 12-month Buy vs Cloud vs API curve
Side-by-Side PK →
Up to 6 models with diff filter, or multi-GPU speed comparison
Popular GPUs & Devices
All 31 →RTX 5090
NVIDIA GeForce RTX 5090 32GB
RTX 5080
NVIDIA GeForce RTX 5080 16GB
RTX 5070 Ti
NVIDIA GeForce RTX 5070 Ti 16GB
RTX 4090
NVIDIA GeForce RTX 4090 24GB
RTX 4070 Ti Super
NVIDIA GeForce RTX 4070 Ti Super 16GB
RTX 4060 Ti 16G
NVIDIA GeForce RTX 4060 Ti 16GB
RTX 4060 8G
NVIDIA GeForce RTX 4060 8GB
RTX 3090
NVIDIA GeForce RTX 3090 24GB