WHAT THIS CARD IS WORTH
ARM Mali-G76 MP12 holds 130 of the models in our catalogue and is, in practice, a Q4_K_M card — the largest it takes is Falcon-H1 7B at Q4_K_M.
TOKENS/SEC PER $100
—
8B at Q4_K_M, so cards compare like for like
VRAM PER $100
—
what memory costs on this card
THE BIGGEST IT TAKES
AGAINST CARDS OF SIMILAR MEMORY
ARM Mali-G76 MP126 GB——
Neighbours in memory rather than in price: memory decides whether a card can do the job at all, so two cards of the same size at different prices is the comparison you are making. Ordered by memory, then bandwidth — not by the value column, which is worked out from bandwidth and price alone and therefore rewards a cheap card whatever its software stack does to that bandwidth in practice. Read it as one input, not as a ranking.
▸ DEVICE UNDER TEST
ARM Mali-G76 MP12 — 6 GB VRAM.
▸ ARM MALI-G76 MP12 SPEC
- BRAND
- Apple
- VRAM
- 6 GB
- BANDWIDTH
- 29.8 GB/s
- FP16 COMPUTE
- 0.4 TFLOPS
- FP32 COMPUTE
- 0.2 TFLOPS
- TDP
- 4 W
- ARCHITECTURE
- Bifrost
▸ AI CAPABILITY
130/ 449 models @ Q4
With 6 GB VRAM and 29.8 GB/s bandwidth, this GPU handles models up to 7B parameters.
Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~3 tok/s.
RENT IT — LIVE PRICES
Checking what the clouds are charging…
§ 01TOP MODELS FOR ARM MALI-G76 MP12
130 FIT · SHOWING 20| MODEL | SIZE | VRAM Q4 | TOK/S | AVG |
|---|---|---|---|---|
| Alpaca 7B | 7B | 4.8 GB | 4 | 27.7 |
| Baichuan2 7B | 7B | 4.8 GB | 4 | 21.5 |
| Vicuna 7B | 7B | 4.8 GB | 4 | 22.0 |
| MPT-7B | 7B | 4.8 GB | 4 | 7.8 |
| Orca 2 7B | 7B | 4.8 GB | 4 | 26.1 |
| WizardLM 2 7B | 7B | 4.8 GB | 4 | 26.1 |
| StarCoder2 7B | 7B | 4.8 GB | 4 | 17.0 |
| WizardCoder Python 7B | 7B | 4.8 GB | 4 | 53.7 |
| WizardLM 7B | 7B | 4.8 GB | 4 | 15.5 |
| OLMo 3.1 RLZero 7B Code | 7B | 4.8 GB | 4 | 21.8 |
| OLMo 3.1 RLZero 7B Math | 7B | 4.8 GB | 4 | 21.8 |
| Dolly v2 7B | 6.9B | 4.7 GB | 4 | 7.0 |
| granite-4.0-h-tiny 6.9B | 6.9B | 4.7 GB | 18 | 49.2 |
| Llama 2 7B | 6.74B | 4.6 GB | 4 | 21.1 |
| CodeLlama 7B | 6.74B | 4.6 GB | 4 | 28.1 |
| LLaMA 1 7B | 6.74B | 4.6 GB | 4 | 30.8 |
| DeepSeek Coder 6.7B | 6.7B | 4.6 GB | 4 | 23.6 |
| OPT 6.7B | 6.7B | 4.6 GB | 4 | 18.5 |
| ChatGLM2 6B | 6.24B | 4.3 GB | 4 | 20.7 |
| ChatGLM3 6B | 6.24B | 4.3 GB | 4 | 42.7 |