Our model, GPU, quantization and API pricing datasets are re-verified on the first weekend of every month. Each entry below lists the date and the source. RSS feed
· Monthly refresh: 2 new models + 1 new GPU, used-market repricing, API price changes
New models: GLM-5.3-Flash (Z.ai, released Aug 25 with September discussion surge — MIT, 321.3B total / 18B active, DSA+KDA hybrid attention, 1M context; config cross-checked via the official HF repo plus unsloth GGUF / llmcheck / OpenRouter) and Qwen3.8-Flash-Next (Alibaba, September — Qwen4 architecture preview, 180B total / 6B active, DeltaNet+QSA hybrid, 256K context; note the Qwen Community License 1.0, not Apache). DeepSeek V4.1-Flash (552B), Kimi K3 (2.8T) and GLM-5.3 flagship (744B) reviewed but excluded: far beyond consumer hardware.
New GPU: Mac Studio M5 Ultra 96GB (announced Aug 25, shipping Sep 22 — 1.2TB/s bandwidth, from $5,499; specs verified via Apple Newsroom / apple.com). No new NVIDIA/AMD cards in September (RTX 50 Super rumored slipped to CES 2027); M5 Max / M5 Pro / M6 machines evaluated but not added pending confirmed specs or pricing.
GPU used prices (US market snapshot, bestvaluegpu 2026-09-30 + GPUDojo/videocardprices/gpupoet September checks): RTX 3090 $1,100→$1,250 (+13.6%), RX 7900 XTX $650→$900 (+38%, memory-shortage rally). RTX 5090 note updated: secondary market $4,899 new / $3,822 used, Amazon listings hit $6,996 (~3.5× MSRP). RTX 4090 ($2,499, +4.1%), RTX 3060, RTX 4070 Ti Super and RTX A6000 checked — no confirmed >10% move.
API pricing: DeepSeek official — deepseek-v4-flash retired Sep 10, replaced by deepseek-flash (V4.1-Flash): off-peak $0.15/$0.60, peak $0.30/$1.20 (verified via benchlm/packet.ai/siliconflow/felloai September reads); V4-Pro peak/off-peak unchanged, the planned Sep-14 switch was cancelled. OpenAI (official help center 9-30 + scriptbyai): added GPT-6.1 Sol $2/$10 (Sep 29), GPT-6 Astra $10/$50 and GPT-6 Luna $0.10/$0.50 (Sep 22); GPT-5.6 Sol's $4/$20 promo is now the de-facto standard rate. Anthropic: Claude Opus 5.5 (Sep 22) at $4/$20 replaces Opus 4.8; Claude Sonnet 5.5 (Sep 28) at $2/$10 matches Sonnet 5; Haiku 4.5 unchanged. Google: Gemini 3.8 Flash added (GA Sep 2, $0.75/$3.75 intro through Dec 31). Together / Groq / DeepInfra: unchanged.
· Monthly refresh: 2 new models, GPU used-market repricing, API price changes
New models: Qwen3.8-27B (Alibaba, Aug 14 — Apache 2.0, hybrid linear/full attention, 262K context; HF config cross-checked via kingy.ai / zenn.dev / NVIDIA developer forums) and Muse Glimmer 30B (Meta, Aug 10 — Apache 2.0, 52 layers, 32Q/2KV GQA, 128K context; config cross-checked via localaimaster / kingy.ai / swfte). Qwen3.8-2.4T-A95B reviewed but excluded: far beyond consumer hardware.
New GPUs: none launched in August 2026 (RTX 50 Super / RX 9080 XT / M5 Ultra Mac Studio remain unannounced).
GPU used prices (US market snapshot, buysellram 2026-08-07/08 + gpupoet/gpudojo): RTX 3060 12GB $180→$260, RTX 3090 $1,000→$1,100, RTX 4070 Ti Super $550→$760, RTX 4090 $1,100→$2,400 (discontinued; priced against RTX 5090). RTX 5090 note updated: secondary market $3,590–4,304 used / $4,288+ new. RTX A6000 and RX 7900 XTX checked, no confirmed >10% move.
API pricing: DeepSeek switched to peak/off-peak pricing on Aug 16 (off-peak V4-Flash $0.22/$0.66, V4-Pro $0.66/$1.98; peak doubles) — api-docs.deepseek.com via codersera/chat-deep.ai. OpenAI (openai.com): GPT-5.6 Terra cut to $2/$12 on Jul 30, GPT-5.6 Luna added at $0.20/$1.20; Sol promo $4/$20 noted. Anthropic cancelled the planned Sep 1 Sonnet 5 increase — $2/$10 is now permanent (Anthropic pricing note, Aug 19). Google: Gemini 3.7 Flash added ($0.75/$3.75 intro through Dec 31). Together: Llama 3.3 70B $0.88→$1.04. Groq / DeepInfra: unchanged.
· Initial dataset verified and published
12 models: specs (params, layers, KV heads, head dim, context) verified against Hugging Face configs; MLA branch confirmed for DeepSeek-R1.
12 GPUs: VRAM, memory bandwidth and MSRP verified against NVIDIA / AMD / Apple official specifications; used-market prices (RTX 3090/4090) sampled from eBay sold listings, July 2026.
API pricing: input/output prices per million tokens verified for DeepSeek, Together AI, Groq, DeepInfra, Fireworks, OpenRouter and Novita (7 providers).
Apple lineup update: 256GB/512GB Mac Studio M3 Ultra configurations discontinued (DRAM shortage); 96GB is the ceiling, base price now $5,299.
GPUFits