Ling 2.6 Flash

inclusionAI · inclusionai/ling-2-6-flash

Provider	inclusionAI
Family	ling
Type	llm-chat
Status	active
Released	2026-04-28
Updated	2026-06-17
Parameters	107.5B
Open weights	yes

What it fits on

Ordered by predicted speed. Bandwidth sets decode rate; memory decides whether it runs at all.

Every figure here is computed, not measured. Fit is weights at each quantisation against device memory, with a 25% allowance for the KV cache, activations and the OS. Decode rate is the memory-bandwidth roofline at 70% efficiency. Nobody has run this model on these devices. This is a mixture-of-experts model and no card carries active-parameter counts, so these predictions use total parameters and understate the real speed.

Device	Memory	Bandwidth	Best quality	Weights	tok/s	Fastest
NVIDIA GB200 Grace Blackwell Superchip	372.0 GB	16000.0 GB/s	bf16	214.99 GB	~52.1	~208.4 q4
AMD Instinct MI355X	288.0 GB	8000.0 GB/s	bf16	214.99 GB	~26.0	~104.2 q4
NVIDIA B200	180.0 GB	7700.0 GB/s	fp8	107.49 GB	~50.1	~100.3 q4
Google TPU7x (Ironwood)	192.0 GB	7380.0 GB/s	fp8	107.49 GB	~48.1	~96.1 q4
AMD Instinct MI325X	256.0 GB	6000.0 GB/s	fp8	107.49 GB	~39.1	~78.1 q4
AMD Instinct MI300X	192.0 GB	5300.0 GB/s	fp8	107.49 GB	~34.5	~69.0 q4
NVIDIA H200 SXM	141.0 GB	4800.0 GB/s	q6	80.62 GB	~41.7	~62.5 q4
NVIDIA H100 NVL	94.0 GB	3900.0 GB/s	q5	67.18 GB	~40.6	~50.8 q4
NVIDIA H100 SXM	80.0 GB	3350.0 GB/s	q4	53.75 GB	~43.6	~43.6 q4
AMD Instinct MI250X	128.0 GB	3200.0 GB/s	q6	80.62 GB	~27.8	~41.7 q4
Google TPU v5p	95.0 GB	2765.0 GB/s	q5	67.18 GB	~28.8	~36.0 q4
NVIDIA A100 80GB SXM	80.0 GB	2039.0 GB/s	q4	53.75 GB	~26.6	~26.6 q4
NVIDIA H100 PCIe	80.0 GB	2000.0 GB/s	q4	53.75 GB	~26.0	~26.0 q4
Apple M3 Ultra	512.0 GB	819.0 GB/s	bf16	214.99 GB	~2.7	~10.7 q4
Apple M2 Ultra	192.0 GB	800.0 GB/s	fp8	107.49 GB	~5.2	~10.4 q4
Apple M4 Max	128.0 GB	546.0 GB/s	q6	80.62 GB	~4.7	~7.1 q4
Apple M2 Max	96.0 GB	400.0 GB/s	q5	67.18 GB	~4.2	~5.2 q4
Apple M3 Max	128.0 GB	400.0 GB/s	q6	80.62 GB	~3.5	~5.2 q4

Where it runs

Taken from the card's availability section.

Platform	Id on that platform
Hugging Face	inclusionAI/Ling-2.6-flash

Verified benchmark evidence

Each row was checked against its source by a reviewer, and carries the model identifier as actually evaluated — which is not always the same as this card's.

Benchmark	Evaluated as	Score	Evidence date	Source kind
aa_lcr	Ling 2.6 Flash	31.33%	2026-09-10 evaluated	independent evaluator	source
critpt	Ling 2.6 Flash	0.0%	2026-09-10 evaluated	independent evaluator	source
gdpval_aa	Ling 2.6 Flash	0.0%	2026-09-10 evaluated	independent evaluator	source
gpqa_diamond	Ling 2.6 Flash	59.29%	2026-09-10 evaluated	independent evaluator	source

Reported benchmark scores

This card reports no benchmark scores yet.

Benchmark	Catalogue standing	Score

Data

This card as JSON · See it in the graph · Edit on GitHub