Topaz Gigapixel | macOS 27 Golden Gate Issue

topazlabs72 · September 21, 2026, 9:49pm

I’ve just updated my M5 Max MacBook Pro to macOS Golden Gate 27.0 (26A428). A clean install of Topaz Gigapixel 1.3.6 for Apple Silicon fails to generate images with Wonder 3.5 / 3 / 2 / 1.

The symptoms are that the Export window says “Error: Error encountered while processing”.

The key part of the Gigapixel log is copied below.

[2026-09-21 14:47:13.115, 287.56 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.114]   
[2026-09-21 14:47:13.115,  70.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.114] [>] Starting upscaling generation... 
[2026-09-21 14:47:13.115,  12.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.114] [>]   Input: 1 frame, 2307x3172px → Padded: 4624x6352px → Output: 4614x6344px (shortest edge: 4614px) 
[2026-09-21 14:47:13.115,  10.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.114] [>]   Batch size: 1, Seed: 42, Channels: RGB 
[2026-09-21 14:47:13.115, 135.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.114]   
[2026-09-21 14:47:13.115,  14.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.114]  ━━━━━━━━ Phase 1: VAE encoding ━━━━━━━━ 
[2026-09-21 14:47:13.116,   1.35 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.116] [A] Encoding batch 1/1 
[2026-09-21 14:47:13.119,   2.70 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.118] [V]   Sequence of 1 frames 
[2026-09-21 14:47:13.134,  15.03 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: [14:47:13.133] [X] [ERROR] Error in Phase 1 (Encoding): coreml_vae: every worker failed to load dec mlpackage within 60.0s; caller should fall back to the MPS path 
[2026-09-21 14:47:13.211,   2.17  s] [1e7efff80] Info | Worker cancel requested 
[2026-09-21 14:47:13.211,  57.00 μs] [1e7efff80] Info | Cancelling processing 
[2026-09-21 14:47:13.211,  15.00 μs] [1e7efff80] Info | [AIE] NeuroserverProcessor::cancel requesting cancel task_id=taskl-efb6b04a94904fc9b1fa79fb1ecee1c7 
[2026-09-21 14:47:13.243, 109.12 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: INFO:topserving.cxctf5t756ug:Received pipe request 
[2026-09-21 14:47:13.277,  33.67 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: Traceback (most recent call last): 
[2026-09-21 14:47:13.277, 140.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "rzyobogmgh6k.py", line 273, in rzyobogmgh6k.Hercules.run 
[2026-09-21 14:47:13.277,  15.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/dcqujnlsy4ni.py", line 1415, in models.bloom_precision.dcqujnlsy4ni.TopazDiffusionDenoiserRamStreaming.__call__ 
[2026-09-21 14:47:13.277,   9.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/dcqujnlsy4ni.py", line 1228, in models.bloom_precision.dcqujnlsy4ni.TopazDiffusionDenoiserRamStreaming._process_frames_core 
[2026-09-21 14:47:13.277,   7.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/njrwkszezfen.py", line 1573, in models.bloom_precision.njrwkszezfen.StreamingBloomPrecisionUpscaler._process_frames_core 
[2026-09-21 14:47:13.277,   7.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/src/core/f97i2mq4q4ay.py", line 562, in models.bloom_precision.src.core.f97i2mq4q4ay.encode_all_batches 
[2026-09-21 14:47:13.277,   7.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/src/core/f97i2mq4q4ay.py", line 513, in models.bloom_precision.src.core.f97i2mq4q4ay.encode_all_batches 
[2026-09-21 14:47:13.277,   6.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "/Applications/Topaz Gigapixel.app/Contents/MacOS/neuroserver/neuroserver/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 124, in decorate_context 
[2026-09-21 14:47:13.277,   6.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:     return func(*args, **kwargs) 
[2026-09-21 14:47:13.277,   6.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:            ^^^^^^^^^^^^^^^^^^^^^ 
[2026-09-21 14:47:13.277,   5.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/src/core/zaf6mzkzb3zz.py", line 244, in models.bloom_precision.src.core.zaf6mzkzb3zz.VideoDiffusionInfer.vae_encode 
[2026-09-21 14:47:13.277,   6.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/src/models/image_vae_2d/b73hddpndy8n.py", line 566, in models.bloom_precision.src.models.image_vae_2d.b73hddpndy8n.ImageAutoencoderKLWrapper.encode 
[2026-09-21 14:47:13.277,   7.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/src/optimization/my6axo8hwp6f.py", line 1602, in models.bloom_precision.src.optimization.my6axo8hwp6f._make_patched_encode.patched_encode 
[2026-09-21 14:47:13.277,   6.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/src/optimization/my6axo8hwp6f.py", line 923, in models.bloom_precision.src.optimization.my6axo8hwp6f._Pool.await_dec_ready 
[2026-09-21 14:47:13.277,   5.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "models/bloom_precision/src/optimization/my6axo8hwp6f.py", line 894, in models.bloom_precision.src.optimization.my6axo8hwp6f._Pool._await_ready_with_quarantine 
[2026-09-21 14:47:13.277,   6.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: RuntimeError: coreml_vae: every worker failed to load dec mlpackage within 60.0s; caller should fall back to the MPS path 
[2026-09-21 14:47:13.277,  76.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: ERROR:local_model_service:Error running video restoration: coreml_vae: every worker failed to load dec mlpackage within 60.0s; caller should fall back to the MPS path 
[2026-09-21 14:47:13.372,  95.25 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: INFO:topserving.cxctf5t756ug:Task taskl-efb6b04a94904fc9b1fa79fb1ecee1c7 progress: 1.0, eta: , message: coreml_vae: every worker failed to load dec mlpackage within 60.0s; caller should fall back to the MPS path 
[2026-09-21 14:47:13.372,  49.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: INFO:topserving.cxctf5t756ug:Task taskl-efb6b04a94904fc9b1fa79fb1ecee1c7 is still running, 60000 seconds left 
[2026-09-21 14:47:13.422,  49.99 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: INFO:topserving.cxctf5t756ug:Pipe request action: cancel 
[2026-09-21 14:47:13.424,   1.98 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: INFO:topserving.cxctf5t756ug:set_cancelled tasks before: {'taskl-8282cc5ae28a482e9eb7c3009f404f93': {'task_id': 'taskl-8282cc5ae28a482e9eb7c3009f404f93', 'status': 'ERROR', 'progress': 0.0, 'eta': '', 'message': '{}', 'is_cancelled': False, 'last_updated_at': 1790027188.495528, 'error_code': 'MODEL_INFERENCE_ERROR'}, 'taskl-8d5908fd54b6402d9acc5eae84dc882c': {'task_id': 'taskl-8d5908fd54b6402d9acc5eae84dc882c', 'status': 'ERROR', 'progress': 0.0, 'eta': '', 'message': '{}', 'is_cancelled': False, 'last_updated_at': 1790027196.110674, 'error_code': 'MODEL_INFERENCE_ERROR'}, 'taskl-efb6b04a94904fc9b1fa79fb1ecee1c7': {'task_id': 'taskl-efb6b04a94904fc9b1fa79fb1ecee1c7', 'status': 'RUNNING', 'progress': 1.0, 'eta': '', 'message': 'coreml_vae: every worker failed to load dec mlpackage within 60.0s; caller should fall back to the MPS path', 'is_cancelled': False, 'last_updated_at': 1790027233.372297, 'error_code': 'OK'}} 
[2026-09-21 14:47:13.424, 213.56 ms] [1e7efff80] Info | [AIE] NeuroserverProcessor::cancel task_id=taskl-efb6b04a94904fc9b1fa79fb1ecee1c7 status=CANCELLED ok=1 
[2026-09-21 14:47:14.378, 953.37 ms] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: INFO:topserving.cxctf5t756ug:Task taskl-efb6b04a94904fc9b1fa79fb1ecee1c7 is RUNNING, 59999 seconds left 
[2026-09-21 14:47:14.378, 230.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: INFO:topserving.cxctf5t756ug:process_message took 2s to complete 
[2026-09-21 14:47:14.378,  36.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: WARNING:topserving.cxctf5t756ug:Error while process_message with task_id: taskl-efb6b04a94904fc9b1fa79fb1ecee1c7 
[2026-09-21 14:47:14.378, 253.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: ERROR:topserving.cxctf5t756ug:ModelInferenceErrorException occurred while processing the message: [MODEL_INFERENCE_ERROR] Error during model inference 
[2026-09-21 14:47:14.379, 593.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: ERROR:topserving.cxctf5t756ug:worker.run An error occurred while processing the message: [MODEL_INFERENCE_ERROR] Error during model inference 
[2026-09-21 14:47:14.379,  61.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: Traceback (most recent call last): 
[2026-09-21 14:47:14.379,  26.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "topserving/cxctf5t756ug.py", line 905, in topserving.cxctf5t756ug.TopServingWorker.run 
[2026-09-21 14:47:14.379,  23.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "topserving/cxctf5t756ug.py", line 810, in topserving.cxctf5t756ug.TopServingWorker.run.receive_and_handle_message 
[2026-09-21 14:47:14.379,  21.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "topserving/cxctf5t756ug.py", line 794, in topserving.cxctf5t756ug.TopServingWorker.run.receive_and_handle_message 
[2026-09-21 14:47:14.379,  20.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]:   File "topserving/d379kb3elyxi.py", line 120, in topserving.d379kb3elyxi.raise_top_serving_exception 
[2026-09-21 14:47:14.379,  19.00 μs] [1726bb000] Info | [AIE] NeuroserverRunner [stderr]: topserving.d379kb3elyxi.ModelInferenceErrorException: [MODEL_INFERENCE_ERROR] Error during model inference

topazlabs72 · September 22, 2026, 12:18am

I should add that other generative models work fine - e.g. Recover 3 and Standard Max all produce output. It just seems to be the Wonder models (the ones I use most) that have this issue.

topazlabs72 · September 22, 2026, 12:54am

…and Gigapixel 1.4.0 beta 2 has the same issue.

alexandre.topazlabs · September 22, 2026, 6:27pm

Can you send us the full logs? We will investigate the issue

brian.masck-0466 · September 22, 2026, 7:33pm

Same issue with MacOS 27 on M5 Pro

2026-09-22-15-26-24.tzlog (327.5 KB)

brian.masck-0466 · September 22, 2026, 11:15pm

marek2 · September 23, 2026, 11:33am

I’m experiencing the same issue; Wonder and Face Recovery don’t work in other modes on macOS. A program error occurs. I have an M5 Pro with 24GB of RAM, 15 CPU cores, and 16 GPU cores.

alexandre.topazlabs · September 23, 2026, 9:32pm

Hello everyone,

Our development team is currently looking into the OS 27 Golden Gate compatibility issue affecting local rendering with Wonder, and Redefine Creative

If you’re running into this, please send us your logs so we can add your case and system information to the development ticket to help facilitate compatibility:

Open Topaz Gigapixel
Go to Help > Open Log Folder.
Send us all of the logs currently in that folder at support@topazlabs.com.
Please mention what issue you have

Please send the logs directly to Support rather than posting them in this forum thread. We’ll review them and relay the relevant information to the development team working on this incompatibility.

In the meantime, Cloud Rendering is free and unlimited, so you can use Cloud Rendering as a workaround.