Name: OpenWeights Terminal measured pairs
Creator: Dvidia
License: https://owterminal.com/cite

49 ranked

Took the throneQwen3.5 2Bvs Sep 15Aug · Qwen3.6-35B-A3B · heldSep · Qwen3.5 2BNations

Liberated

#	Model · Build	Hardware			Runtime	Proof
1NEW	Qwen3.5 2BQ4_K_M RTX 5090	RTX 5090	351	—	—	@fillagrew
2NEW	Qwen3.8-27BNVFP4 RTX PRO 6000 Blackwell Max-Q	RTX PRO 6000 Blackwell Max-Q	272.8	—	TensorFold	@llm_frontier_jp
3NEW	GLM 5.3 FlashEXL3 4× RTX PRO 6000	4× RTX PRO 6000	264	—	TensorFold 0.6.2	@Kurcide
4NEW	Swift 1.5 Qwen3.8 Flash-NextIQ2_XS RTX 4090	RTX 4090	220	3830	Strata	@Knuckles_XBT
5NEW	Qwen3.8-27BNVFP4 RTX A5000	RTX A5000	177.6	—	TensorFold	@Ja6ek
6NEW	Gemma-4 26B-A4B-itQ4_K_M RTX 5090	RTX 5090	173	—	—	@fillagrew
7NEW	Qwen3.8-27BNVFP4 RTX PRO 6000 Blackwell Max-Q	RTX PRO 6000 Blackwell Max-Q	151.4	—	vLLM	@llm_frontier_jp
8NEW	Qwen3.8-Flash-NextIQ3_S RTX 5090	RTX 5090	142	—	Strata	@arianpg
9NEW	Qwen3.8-27BNVFP4 RTX A5000	RTX A5000	139.3	—	vLLM	@Ja6ek
10NEW	Qwen3.8-Flash-Next M5 Max	M5 Max	126.5	—	MTPLX V2.11.3	@Youssofal_
11NEW	GLM 5.3 FlashEXL3 4bpw RTX PRO 6000 + 2× GB10	RTX PRO 6000 + 2× GB10	120	—	TensorFold	@hokazuya
12▼11	Qwen3.6-35B-A3B4-bit M4 Pro 64GB	M4 Pro 64GB	108.6	—	Rapid-MLX	@draslan_eth
13NEW	Qwen3.8-Flash-NextIQ3_S RTX 5090	RTX 5090	104	1657	Strata	@DaSun64125381
14NEW	Qwen3.8-27B4-bit DGX Spark	DGX Spark	102.9	—	TensorFold	@Oluwaphilemon1
15NEW	Qwen3.8-Flash-NextIQ2_XS RTX 3090 24GB	RTX 3090 24GB	100.6	—	Strata	@draslan_eth
16▼14	Qwen3.6-35B-A3BNVFP4 DGX Spark	DGX Spark	94.4	181.5	—	@bonellisystems
17NEW	Qwen3.8-27B4-bit M5 Pro 64GB	M5 Pro 64GB	89.9	—	TensorFold	@aartiles24
18▼15	Qwen3.6-35B-A3B4bit M4 Pro	M4 Pro	85.5	—	rapid-mlx	@rapidmlx
19NEW	Qwen3.5-4B4bit M4 Pro	M4 Pro	82.8	—	rapid-mlx	@rapidmlx
20NEW	Qwen3.8-Flash-NextNVFP4 RTX 5090	RTX 5090	80.9	—	—	@tekizaihq
21NEW	Qwen3.8-Flash-NextEXL3 DGX Spark	DGX Spark	79.5	—	EXL3	@yume_arasaki
22NEW	Qwen3.8-Flash-NextNVFP4 DGX Spark	DGX Spark	74.8	—	TensorFold	@redp314
23NEW	Qwen3.8-27BNVFP4 DGX Spark	DGX Spark	72	—	—	@hasso5703
24NEW	Qwen3.8-Flash-NextIQ3_S RTX 3090 24GB	RTX 3090 24GB	72	—	Strata	@needmorevram
25NEW	AliceAI-Foundation-80B-A3BQ4_K_M M5 Max	M5 Max	65	—	llama.cpp	@aqty
26NEW	Qwen3.8-27B4-bit M6 mini 32GB	M6 mini 32GB	61.9	—	TensorFold	@jmurillocode
27NEW	GLM 5.3 FlashEXL3 DGX Spark	DGX Spark	60.4	1950	TensorFold	@MiaAI_lab
28NEW	GLM 5.3 FlashNVFP4 DGX Spark	DGX Spark	55.6	—	—	@PlusTen_AI
29NEW	Qwen3.8-Flash-NextIQ3_S RTX 3080 10GB	RTX 3080 10GB	52.7	—	Strata	@needmorevram
30NEW	Empero Qwen3.8-35B-A3BQ4_K_M RTX 3060	RTX 3060	50	550	llama.cpp	@Oluwaphilemon1
31NEW	Qwen3.8-Flash-NextIQ3_S RTX 4060 Ti	RTX 4060 Ti	50	—	Strata	@xiaobei1943
32NEW	Qwen3.5-9B4bit M4 Pro	M4 Pro	49.3	—	rapid-mlx	@rapidmlx
33▼29	Qwen3-8B4bit M4 Pro	M4 Pro	48.3	—	rapid-mlx	@rapidmlx
34NEW	MiMo-V2.6-Distill-Qwen-9BQ5_K_M RTX 5060	RTX 5060	47	1600	llama.cpp	@stfu0911
35NEW	Qwen3.8-Flash-NextFP8 DGX Spark	DGX Spark	45	—	—	@Oluwaphilemon1
36NEW	Qwen3.8-27BNVFP4 DGX Spark	DGX Spark	42.3	—	TensorFold 0.6.3	@edurdias
37NEW	Qwen3.8-27BUD-Q4_K_XL RTX 4090	RTX 4090	40.7	2659.8	llama.cpp	@Oluwaphilemon1
38NEW	Ornith GTX 1660 SUPER	GTX 1660 SUPER	40	—	—	@mine_craft_bui
39▼34	Qwen3.6-35B-A3BUD-IQ3_XXS M1 Max 32GB	M1 Max 32GB	37	350	—	@moonsteroid
40NEW	Qwen3.8-27BNVFP4 DGX Spark	DGX Spark	31.8	—	SGLang 0.5.21	@edurdias
41NEW	GLM 5.3Int4/Int8 DGX Spark	DGX Spark	30	840	sparkDash	@majewskizby
42NEW	Qwen3.8-27BNVFP4 DGX Spark	DGX Spark	28.7	—	vLLM 0.30.0	@edurdias
43NEW	Qwen3.8-Flash-NextIQ2_XS RTX 3060	RTX 3060	27.7	—	Strata	@dec21ai
44NEW	Qwen3.8-27BIQ3_XXS RTX 5060 Ti	RTX 5060 Ti	23	—	llama.cpp	@fntAInhead
45NEW	Qwen3.8-27BUD-Q2_K_XL RTX 5060 Ti	RTX 5060 Ti	19	—	llama.cpp	@fntAInhead
46NEW	GLM 5.3 FlashEXL3 DGX Spark	DGX Spark	18.6	—	EXL3	@vcruz305
47NEW	GLM 5.3 FlashNVFP4 DGX Spark	DGX Spark	15	—	vLLM	@sudoingX
48NEW	Nex-N2.5-miniQ4_K_M RTX 5060 Ti	RTX 5060 Ti	14	—	llama.cpp	@fntAInhead
49NEW	Qwen3.8-27B TurboFCFusionIQ2_M RTX 5060 Ti	RTX 5060 Ti	10	—	llama.cpp	@fntAInhead

Qwen3.6-35B-A3B

M4 Pro 64GB

M4 Pro 64GB · Rapid-MLX · 4-bit

Pair

#12 / 49 in this filter

#1 / 1 on this chip

#1 / 3 of this model

Fastest of this model

Decode · tok/s

Prefill

Peak

This model elsewhere

Lineage

Officialclaimed

Originating sentence

Qwen3.6-35B-A3B (4-bit, ~3B active, Rapid-MLX) vs. Nemotron 3.5 Lightning 30B-A3B (4-bit, ~3–3.6B active, TensorFold) on M4 Pro 64GB. Qwen: 108.6 tok/s + 15/15 tools.

Flags

~3B active · 15/15 tools · 5/5 needles at 245K

Caveat

First-person M4 Pro 64GB. Rank is the quoted 108.6. Nemotron side has no tok/s in the quote.

Proof

@draslan_eth

Get

Get unlinked

Pairs

Liberated