Ling-mini-2.0

inclusionAI · inclusionai/ling-mini-2-0

ProviderinclusionAI
Familyling
Typellm-chat
Statusactive
Released2025-09-08
Updated2026-04-13
Parameters16.3B
Open weightsyes

Lineage

RelationshipModel
Is a finetune ofinclusionAI/Ling-mini-base-2.0

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.
DeviceMemoryBandwidthBest qualityWeightstok/sFastest
Cerebras WSE-344.0 GB21000000.0 GB/sbf1632.51 GB~5129532.5~20518130.2 q4
NVIDIA GB200 Grace Blackwell Superchip372.0 GB16000.0 GB/sbf1632.51 GB~3908.2~15632.9 q4
AMD Instinct MI355X288.0 GB8000.0 GB/sbf1632.51 GB~1954.1~7816.4 q4
NVIDIA B200180.0 GB7700.0 GB/sbf1632.51 GB~1880.8~7523.3 q4
Google TPU7x (Ironwood)192.0 GB7380.0 GB/sbf1632.51 GB~1802.7~7210.7 q4
AMD Instinct MI325X256.0 GB6000.0 GB/sbf1632.51 GB~1465.6~5862.3 q4
AMD Instinct MI300X192.0 GB5300.0 GB/sbf1632.51 GB~1294.6~5178.4 q4
NVIDIA H200 SXM141.0 GB4800.0 GB/sbf1632.51 GB~1172.5~4689.9 q4
NVIDIA H100 NVL94.0 GB3900.0 GB/sbf1632.51 GB~952.6~3810.5 q4
NVIDIA H100 SXM80.0 GB3350.0 GB/sbf1632.51 GB~818.3~3273.1 q4
AMD Instinct MI250X128.0 GB3200.0 GB/sbf1632.51 GB~781.6~3126.6 q4
Google TPU v5p95.0 GB2765.0 GB/sbf1632.51 GB~675.4~2701.6 q4
NVIDIA A100 80GB SXM80.0 GB2039.0 GB/sbf1632.51 GB~498.1~1992.2 q4
NVIDIA H100 PCIe80.0 GB2000.0 GB/sbf1632.51 GB~488.5~1954.1 q4
NVIDIA GeForce RTX 509032.0 GB1792.0 GB/sfp816.26 GB~875.4~1750.9 q4
Google TPU v6e (Trillium)32.0 GB1638.0 GB/sfp816.26 GB~800.2~1600.4 q4
AMD Instinct MI21064.0 GB1600.0 GB/sbf1632.51 GB~390.8~1563.3 q4
NVIDIA A100 40GB SXM40.0 GB1555.0 GB/sfp816.26 GB~759.7~1519.3 q4
Google TPU v432.0 GB1200.0 GB/sfp816.26 GB~586.2~1172.5 q4
NVIDIA GeForce RTX 409024.0 GB1008.0 GB/sfp816.26 GB~492.4~984.9 q4
AMD Radeon RX 7900 XTX24.0 GB960.0 GB/sfp816.26 GB~469.0~938.0 q4
NVIDIA GeForce RTX 508016.0 GB960.0 GB/sq510.16 GB~750.4~938.0 q4
NVIDIA GeForce RTX 309024.0 GB936.0 GB/sfp816.26 GB~457.3~914.5 q4
NVIDIA L40S48.0 GB864.0 GB/sbf1632.51 GB~211.0~844.2 q4
Apple M3 Ultra512.0 GB819.0 GB/sbf1632.51 GB~200.1~800.2 q4
AMD Radeon RX 7900 XT20.0 GB800.0 GB/sq612.19 GB~521.1~781.6 q4
Apple M2 Ultra192.0 GB800.0 GB/sbf1632.51 GB~195.4~781.6 q4
Google TPU v5e16.0 GB800.0 GB/sq510.16 GB~625.3~781.6 q4
NVIDIA GeForce RTX 4080 SUPER16.0 GB736.0 GB/sq510.16 GB~575.3~719.1 q4
NVIDIA GeForce RTX 4070 Ti SUPER16.0 GB672.0 GB/sq510.16 GB~525.3~656.6 q4
AMD Radeon RX 9070 XT16.0 GB640.0 GB/sq510.16 GB~500.3~625.3 q4
Apple M4 Max128.0 GB546.0 GB/sbf1632.51 GB~133.4~533.5 q4
Apple M1 Max64.0 GB400.0 GB/sbf1632.51 GB~97.7~390.8 q4
Apple M2 Max96.0 GB400.0 GB/sbf1632.51 GB~97.7~390.8 q4
Apple M3 Max128.0 GB400.0 GB/sbf1632.51 GB~97.7~390.8 q4
NVIDIA GeForce RTX 3060 12GB12.0 GB360.0 GB/sq48.13 GB~351.7~351.7 q4
NVIDIA L424.0 GB300.0 GB/sfp816.26 GB~146.6~293.1 q4
NVIDIA GeForce RTX 4060 Ti 16GB16.0 GB288.0 GB/sq510.16 GB~225.1~281.4 q4
Apple M4 Pro64.0 GB273.0 GB/sbf1632.51 GB~66.7~266.7 q4
Apple M432.0 GB120.0 GB/sfp816.26 GB~58.6~117.2 q4

Where it runs

Taken from the card's availability section.

PlatformId on that platform
Hugging FaceinclusionAI/Ling-mini-2.0

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.
BenchmarkEvaluated asScoreEvidence dateSource kind
critptLing-mini-2.00.0%2026-09-10 evaluatedindependent evaluatorsource
gpqa_diamondLing-mini-2.056.16%2026-09-10 evaluatedindependent evaluatorsource

Reported benchmark scores

This card reports no benchmark scores yet.

BenchmarkCatalogue standingScore

Data

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