Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchComfyUI_NAIDGenerator lists the NovelAI V5 model IDs nai-diffusion-5-full and nai-diffusion-5-curated, and a developer reported successful image-generation responses from both. The separate “Error fetching Anlas” 400 reported in the same test came from the extension’s balance lookup—not from either model request. The author traced that lookup to a user-data endpoint and reported a fix, but the endpoint details are an individual test, not an official NovelAI API guarantee.
Can ComfyUI run NovelAI V5?
The ComfyUI_NAIDGenerator README lists nai-diffusion-5-full and nai-diffusion-5-curated among its V4+ model identifiers. The README initially cautioned that the V5 identifiers had not been independently confirmed against the live API. In a later report, Ilan Kim said both IDs returned successful responses through the node in a test using ComfyUI 0.35.0 and ComfyUI_NAIDGenerator version 1.0.9, pinned to commit 52da75a. That is evidence they worked in that setup, not a guarantee for every extension revision or account. Read the developer’s test report and check the extension README for the project’s current model list and instructions.
As an Amazon Associate I earn from qualifying purchases.
Which V5 model should you choose?
NovelAI describes Full as trained on a more comprehensive, less curated dataset, and suggests trying it when Curated does not produce a specific desired result. Curated uses a smaller set of cleaner, more focused images and is presented as a general-use option. These are vendor descriptions, not a guarantee that one model will work better for every prompt.
| Model | NovelAI’s description | Effective prompt-token guidance | Size in the reported ComfyUI test |
|---|---|---|---|
nai-diffusion-5-full |
More comprehensive, less curated training set; try it for a specific result Curated does not produce. | Up to approximately 1,471 effective tokens over the base prompt. | 832×1216 in Ilan Kim’s test. |
nai-diffusion-5-curated |
Smaller training set of cleaner, more focused images; described as a general-use choice. | Up to approximately 703 effective tokens over the base prompt. | 512×768 in Ilan Kim’s test. |
The token figures are from NovelAI’s model documentation; the image sizes are the author’s test settings, not recommended or maximum resolutions. NovelAI Image Models documentation describes the model differences. NovelAI says V5 is available to subscribing tiers and identifies it as its first model generation with a usage limit on free Opus generations. Availability or cost for a particular account should not be inferred from another user’s test.
#1 Best Overall
- Powered by Radeon RX 9070 XT
- WINDFORCE Cooling System
- Hawk Fan
- Server-grade Thermal Conductive Gel
- RGB Lighting
What does “Error fetching Anlas” 400 mean?
It may indicate that the extension cannot retrieve your account balance, even when image generation works. In Kim’s report, the console message was “Failed to fetch Anlas balance: 400 Client Error: Bad Request.” The author traced the balance request to https://api.novelai.net/user/data; an authenticated comparison reportedly returned 400 from that host and 200 from https://image.novelai.net/user/data.
Changing the host was only part of the reported fix. The response from the latter endpoint represented subscription.trainingStepsLeft as an object containing fixedTrainingStepsLeft and purchasedTrainingSteps, rather than as a single integer. Code that expects one integer may still display the balance incorrectly or fail when calculating the difference before and after generation. The report says the extension code needed to normalize that object-shaped value as well as use the other host. NovelAI’s reviewed public model documentation does not establish this endpoint or response format as an official API contract, so treat the details as a developer’s observed workaround rather than a guaranteed permanent fix.
Rank #2
- NVIDIA Ampere Streaming Multiprocessors: The all-new Ampere SM brings 2X the FP32 throughput and improved power efficiency.
- 2nd Generation RT Cores: Experience 2X the throughput of 1st gen RT Cores, plus concurrent RT and shading for a whole new level of ray-tracing performance.
- 3rd Generation Tensor Cores: Get up to 2X the throughput with structural sparsity and advanced AI algorithms such as DLSS. These cores deliver a massive boost in game performance and all-new AI capabilities.
- Axial-tech fan design features a smaller fan hub that facilitates longer blades and a barrier ring that increases downward air pressure.
- OC Mode : 1500 MHz (Boost Clock)/Default Mode : 1470 MHz (Boost Clock)
How to troubleshoot the error without confusing balance tracking and generation
- Identify which operation failed. Check whether the message concerns the Anlas Tracker or whether the image-generation request itself returned an error. A balance lookup can fail independently of generation.
- Check your extension revision. Compare the installed ComfyUI_NAIDGenerator version or commit with the version under which the issue was reported: 1.0.9, commit
52da75a. Later or earlier code may behave differently. - If you maintain a fork, inspect both balance-handling details. The reported change switched the user-data host to
image.novelai.netand handled the object-shapedtrainingStepsLeftvalue. This is a report-specific implementation clue, not an official API specification. - Test generation separately from balance display. Kim reported that both V5 IDs generated successfully while the tracker’s old balance request returned 400. A working image request does not prove the balance tracker is fixed, and a broken tracker does not by itself prove V5 generation is broken.
- Check timeouts if a request appears to stall. The extension README warns that a timeout set too short may prevent the image from being received even if Anlas is charged. Its README is the relevant place to review the setting and current usage guidance.
The report also mentions a 400 during generation on a trial account, but the author did not reproduce it or establish that it shared the balance-lookup cause. The report did not include a V4.5 regression test after the balance-code change. If generation itself fails, investigate that request as a separate issue rather than applying the balance diagnosis automatically.
What the reported test does—and does not—show
In the author’s configuration, nai-diffusion-5-full at 832×1216 returned a 200 response in 4.1 seconds, while nai-diffusion-5-curated at 512×768 returned a 200 response in 1.6 seconds. The author also said both PNGs contained source metadata identifying NovelAI Diffusion V5. These are observations from one setup, not performance benchmarks or compatibility guarantees.
Rank #3
- Powered by the NVIDIA Blackwell architecture and DLSS 4. System Requirements: Minimum 850W PSU with 16-pin 12V-2x6 (12VHPWR) connector required. Verify before purchasing.
- Military-grade components deliver rock-solid power and longer lifespan for ultimate durability. Compatibility: 348mm (13.7") length, 3.6 slots, 4.3 lbs. Confirm case clearance and slot spacing. GPU bracket included.
- Protective PCB coating helps protect against short circuits caused by moisture, dust, or debris
- 3.6-slot design with massive fin array optimized for airflow from three Axial-tech fans
- Phase-change GPU thermal pad helps ensure optimal thermal performance and longevity, outlasting traditional thermal paste for graphics cards under heavy loads
The same report recorded 0 Anlas for one Normal-size 832×1216 image at 28 steps on an Opus account. That single account-and-configuration result does not establish that V5 images are always free or that other accounts will be charged the same way. NovelAI’s model page also says V5 was trained in-house using 268,000 B200 GPU-hours; that is the company’s statement, not an independent measurement.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
