Will It Run AI
CalculatorModelsHardwareCompare
Product
  • Calculator
  • Compare
  • Tier List
Browse
  • Models
  • Hardware
  • Docs
About
  • Why It Works
  • What's New
  • Legal Notice
  • Privacy Policy

All estimates are approximations based on mathematical models and public specifications. Actual performance may vary. Do not make purchasing decisions based solely on these estimates.

Data sourced from Hugging Face, Ollama, and official model documentation. Model names and logos are trademarks of their respective owners.

© 2026 Will It Run AI — Fase Consulting Ibiza, S.L. (NIF: B57969656)

Home/Hardware/GPUs/Intel Arc Pro B60 24GB
Intel

Intel

Intel Arc Pro B60 24GB

Arc Pro BWorkstationBattlemagePCIe 5oneAPI
24GB
VRAM
456GB/s
Bandwidth
12TFLOPS
FP16 Compute
197TOPS
INT8 Inference
VRAM24 GBBandwidth456 GB/sCompute12 TFInference197 TOPS
Intel Arc Pro B60 24GBCategory AvgAMD Instinct MI100 32GB

Specifications

Compute
FP1612.279999732971191 TFLOPS
INT8197 TOPS
ArchitectureBattlemage
Memory
VRAM24 GB
Bandwidth456 GB/s
General
FamilyArc Pro B
SegmentWorkstation
InterconnectPCIe 5
Compute PlatformONEAPI

Architecture

Battlemage

Battlemage is Intel's second-generation Arc GPU architecture (Xe2-HPG), built on TSMC N4. It delivers significant performance-per-watt improvements over Alchemist with enhanced XMX engines and improved driver maturity.

AI Relevance

Better driver stability and improved XMX throughput make Battlemage more viable for AI inference than Alchemist. The Arc B580 (12 GB) is an increasingly popular budget option for local LLM experimentation via SYCL/oneAPI backends in llama.cpp.

Process: TSMC N4Platform: ONEAPIPrecisions: FP32, FP16, BF16, INT8

Recommendations by Workload

Agentic Coding

C

Qwen 2.5 Coder 14B

This model is still usable for agentic-coding, but it is not the most specialized pick. It sits in the middle of the current model mix. It fits natively with comfortable headroom. Known channels: huggingface, ollama, lm-studio.

Decode 28.8 tok/s · 47K ctx · llama.cpp
16.2 GB / 24.0 GB VRAM

Chat

C

Qwen 3 14B

This model is a direct match for chat. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, ollama, lm-studio.

Decode 28.8 tok/s · 15K ctx · llama.cpp
12.9 GB / 24.0 GB VRAM

Coding

C

Devstral Small 2 24B Instruct

This model is a direct match for coding. It belongs to a current frontier family for local AI. It should run, but memory headroom will be limited. Known channels: huggingface, ollama, lm-studio.

Decode 16.8 tok/s · 18K ctx · llama.cpp
21.7 GB / 24.0 GB VRAM

RAG

C

granite 8b code instruct 4k

This model is a direct match for rag. It sits in the middle of the current model mix. It fits natively with comfortable headroom.

Decode 50.5 tok/s · 72K ctx · llama.cpp
10.7 GB / 24.0 GB VRAM

Reasoning

C

Qwen 3 14B

This model is a direct match for reasoning. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, ollama, lm-studio.

Decode 28.8 tok/s · 27K ctx · llama.cpp
14.0 GB / 24.0 GB VRAM

Full Model Compatibility

AlibabaQwen3-VL 30B A3B Instruct
C51
30B22.4 GB35 tok/s17K ctx
moe
AlibabaQwen3-Coder 30B A3B Instruct
C51
30.5B22.7 GB34 tok/s17K ctx
moe
UnslothQwen3.5 9B
C49
9B10.2 GB45 tok/s38K ctx
dense
HauhauCSHQwen3.5 9B Uncensored HauhauCS Aggressive
C49
9B10.2 GB45 tok/s38K ctx
dense
Lmstudio-communityLQwen3.5 9B
C49
9B10.2 GB45 tok/s38K ctx
dense
MistralCodestral 2 25.08
C49
22B20.2 GB18 tok/s19K ctx
dense
MistralDevstral Small 2 24B Instruct
C49
24B21.7 GB17 tok/s18K ctx
dense
BartowskiBMeta Llama 3.1 8B Instruct
C49
8B9.4 GB51 tok/s41K ctx
dense
XtunerXllava llama 3 8b v1 1
C49
8B9.4 GB51 tok/s41K ctx
dense
MistralDevstral Small 1.1
C49
24B21.7 GB17 tok/s18K ctx
dense
UnslothDeepSeek R1 0528 Qwen3 8B
C49
8B9.4 GB51 tok/s41K ctx
dense
MaziyarPanahiMMeta Llama 3 8B Instruct
C49
8B9.4 GB51 tok/s41K ctx
dense
TheBlokeTLlama 2 7B Chat
C48
7B8.7 GB58 tok/s44K ctx
dense
UnslothQwen3.5 27B
C48
27B24.0 GB15 tok/s16K ctx
dense
TheBlokeTMistral 7B Instruct v0.2
C48
7B8.7 GB58 tok/s44K ctx
dense
MaziyarPanahiMMistral 7B Instruct v0.3
C48
7B8.7 GB58 tok/s44K ctx
dense
Unslothgemma 3 27b it
C48
27B24.0 GB15 tok/s16K ctx
dense
UnslothQwen3.5 4B
C48
4B6.5 GB101 tok/s59K ctx
dense
Lmstudio-communityLgemma 3 4b it
C48
4B6.5 GB101 tok/s59K ctx
dense
BartowskiBLlama 3.2 3B Instruct
C48
3B6.3 GB116 tok/s61K ctx
dense
QwenQwen2.5 3B Instruct
C47
3B5.9 GB135 tok/s65K ctx
dense
BartowskiBgemma 2 2b it
C47
2B5.7 GB158 tok/s67K ctx
dense
Googlegemma 2b
C47
2B5.3 GB202 tok/s72K ctx
dense
TheDrummerTGemmasutra Mini 2B v1
C47
2B5.3 GB202 tok/s72K ctx
dense
QwenQwen2.5 1.5B Instruct
C47
1.5B5.0 GB246 tok/s77K ctx
dense
Hugging-quantsHLlama 3.2 1B Instruct Q8 0
C47
1B4.9 GB259 tok/s78K ctx
dense
TheBlokeTTinyLlama 1.1B Chat v1.0
C46
1.1B4.8 GB246 tok/s80K ctx
dense
Ggml-orgGSmolVLM 500M Instruct
C46
0.5B4.5 GB259 tok/s85K ctx
dense
Ggml-orgGembeddinggemma 300M
C46
0.3B4.3 GB259 tok/s88K ctx
dense
AlibabaQwen 2.5 Coder 32B
D37
32B27.8 GB11 tok/s14K ctx
dense
DeepSeekDeepSeek R1 671B
F0
671B418.4 GB2 tok/s4K ctx
moe
MistralDevstral 2 123B Instruct
F0
123B97.5 GB3 tok/s4K ctx
dense
Z.aiGLM-5
F0
744B463.4 GB2 tok/s4K ctx
moe
UnslothQwen3.5 35B A3B
F0
35B30.1 GB12 tok/s13K ctx
dense
Moonshot AIKimi K2.5
F0
1000B618.3 GB2 tok/s4K ctx
moe
MistralMistral Large 3
F0
675B421.5 GB2 tok/s4K ctx
+1moe
MistralMistral Small 4 119B
F0
119B76.9 GB10 tok/s5K ctx
moe
AlibabaQwen3-Coder 480B A35B Instruct
F0
480B301.6 GB2 tok/s4K ctx
moe
AlibabaQwen3-Coder-Next
F0
80B52.9 GB15 tok/s7K ctx
moe
UnslothQwen3.5 122B A10B
F0
122B82.1 GB4 tok/s5K ctx
dense
DeepSeekDeepSeek V3 671B
F0
671B418.4 GB2 tok/s4K ctx
moe
MistralMixtral 8x22B
F0
141B95.4 GB6 tok/s4K ctx
moe
AlibabaQwen 2.5 72B
F0
72B58.5 GB6 tok/s7K ctx
dense
AlibabaQwen 3 235B A22B
F0
235B150.1 GB5 tok/s4K ctx
moe
UnslothQwen3.5 397B A17B
F0
397B307.5 GB2 tok/s4K ctx
dense
MetaLlama 3.3 70B
F0
70B56.9 GB6 tok/s7K ctx
dense
MetaLlama 4 Maverick 17B 128E
F0
400B250.0 GB3 tok/s4K ctx
moe
CohereCommand A 111B
F0
111B88.4 GB4 tok/s4K ctx
dense
AlibabaQwen 2.5 VL 72B
F0
72B58.5 GB6 tok/s7K ctx
dense
Lmstudio-communityLQwen3.5 35B A3B
F0
35B30.1 GB12 tok/s13K ctx
dense

Just out of reach

Models you could run with an upgrade

High-quality models that need a bit more memory

DeepSeekDeepSeek R1 671B
671BTier 5Needs ~424.2 GB
MistralDevstral 2 123B Instruct
123BTier 5Needs ~116.8 GB
Runs on Mac Studio M3 Ultra 256GB
Z.aiGLM-5
744BTier 5Needs ~469.6 GB
UnslothQwen3.5 35B A3B
35BTier 5Needs ~35.6 GB
Runs on Mac mini M4 64GB (~$1,099)
Moonshot AIKimi K2.5
1000BTier 5Needs ~623.3 GB

Upgrade paths

Upgrade from Intel Arc Pro B60 24GB

See what you unlock with more powerful hardware

Upgrade options

Upgrade options

AMDAMD Instinct MI100 32GBNext step up
32 GB VRAM (+8)1228 GB/s (+772)
A
Unlocks Qwen3.5 35B A3B, Qwen 2.5 Coder 32B, Qwen3.5 35B A3B+14 more · +208% faster avg

 

AppleMac mini M4 64GBBest value
64 GB Unified (+40)
B
Unlocks Qwen3.5 35B A3B, Qwen 2.5 Coder 32B, Qwen3.5 35B A3B+16 more

~$1,099 MSRP

IntelIntel Data Center GPU Max 1550 128GBIntel upgrade
128 GB VRAM (+104)3200 GB/s (+2744)
A
Unlocks Devstral 2 123B Instruct, Qwen3.5 35B A3B, Mistral Small 4 119B+38 more · +624% faster avg

 

AMDAMD Instinct MI350X 288GBBiggest leap
288 GB VRAM (+264)8000 GB/s (+7544)
A
Unlocks Devstral 2 123B Instruct, Qwen3.5 35B A3B, Mistral Small 4 119B+46 more · +1947% faster avg

~$8,000 MSRP

Compare this GPU