Short answer: There is no single “best” SBC for Teams Direct Routing. AudioCodes, Ribbon, and anynode (TE-SYSTEMS) are all Microsoft-certified with media bypass and Local Media Optimization support. Pick by what you already run, who will support it at 2am, and whether you need appliance scale or a software-first footprint. Wrong fit shows up as cert/TLS fights, Inactive OPTIONS, or 488s — not a fancy feature matrix on a slide.

What “best” actually means on a ticket

You’ll see RFPs ask “which SBC should we buy for Teams?” like there’s a winner. There isn’t. Microsoft only supports Phone System with Direct Routing on certified SBC firmware. All three vendors on this page sit on that list with non-media-bypass, media bypass, and (on supported versions) Local Media Optimization.

So the decision is operational:

• Who owns the box when OPTIONS go Inactive?
• Do you already have AudioCodes Mediant gear, Ribbon Edge/Core, or a software SBC culture?
• Branch edge vs carrier core vs small cloud VM?
• Do you need ELIN / emergency quirks your region cares about?

Don’t start with marketing decks. Start with your SIP trunk provider’s known-good guide for that vendor, your cert story, and who answers Sev-1.

Quick compare (practitioner view)

Shape	Hardware Mediant line + Virtual / Cloud Edition	SBC Edge (1000/2000/SWe Edge) + Core (5400/7000/SWe)	Software SBC; wizard-heavy Teams templates
Teams Direct Routing	Certified; huge install base in enterprise DR	Certified; strong in carrier / large edge estates	Certified; popular for lean software deployments
Media bypass	Yes (supported builds)	Yes (supported builds)	Yes (supported builds)
Local Media Optimization	Yes on listed Mediant / VE / CE builds	Yes on listed Edge and Core builds	Yes from 4.0.1+
ELIN capable (MS table)	Yes on listed products	Yes on Edge family; Core rows vary — check the current Learn table	Yes
Typical feel	INI / Web UI, deep SIP manipulation, lots of field notes	Ribbon UX; Edge vs Core skill split matters	Guided Teams + SIP provider wizards; less “appliance ritual”
You’ll like it if…	You already backup Mediant configs and know IP Groups / Proxy Sets	You’re standardizing on Ribbon for trunks and enterprise edge	You want software on existing VMs / cloud and fast DR standup
Watch-outs	Config sprawl; version pin to certified major.minor	Picking Edge vs Core wrong for the use case	Ops maturity / HA design still on you — software ≠ “set and forget”

Version numbers move. Always re-check Microsoft’s certified SBC table before you freeze a BOM. Firmware higher than documented is supported when the major.minor matches the certified line — confirm with the vendor if you’re unsure.

AudioCodes: the enterprise default you’ll meet most

If you’ve done Direct Routing for more than a year, you’ve probably touched a Mediant. Hardware from Mediant 500/800 up through 4000/7000/9000, plus Virtual Edition and Cloud Edition — all on Microsoft’s certified list with media bypass and LMO on the supported trains (commonly the 7.40A.x family for current DR).

Why shops stick with it:

• Huge community of INI examples and TAC muscle memory
• Fine-grained SIP message manipulation when carriers send weird PAI / From / Contact
• Same family for small branch boxes and bigger cores

What bites people: treating every Mediant like a unique snowflake. Standardize Proxy Sets, IP Groups, TLS contexts, and a backup habit. A Mediant 4000 without a recent backup is a future outage — see AudioCodes Mediant 4000 backup and restore. For a Teams-oriented walkthrough on this site, use configure AudioCodes for Direct Routing and the architecture notes in Direct Routing SBC configuration.

Ribbon: Edge vs Core — don’t mix the stories

Ribbon shows up twice in people’s heads: the Edge line (SBC 1000, 2000, SWe Edge / former SWe Lite, Edge 8000 with SWe Edge) and the Core line (5400, 7000, SWe, hosted variants). Both are certified for Direct Routing with media bypass; LMO sits on the builds Microsoft lists for each family.

You’ll see Ribbon when:

• The enterprise already standardized Edge appliances at sites
• A carrier / MSP runs Core or SWe at scale
• You’re replacing older Sonus / EdgeMarc footprints (EdgeMarc is on the DR table separately — treat it as its own product row)

Don’t start with “Ribbon is the carrier SBC” as a slogan. Ask which product code is in the rack. Edge ops and Core ops are different skill sets. Mis-sizing is how you end up with an Edge box drowning in concurrent sessions or a Core design nobody on the Teams team can touch.

anynode: software-first, wizard-friendly

anynode from TE-SYSTEMS is a full software SBC. Microsoft lists it with media bypass, LMO (4.0.1+), ELIN, and 911 provider capability on the current certification table. Recommended train in the public list has been in the 4.x range — pin what your partner actually ships and keep major.minor aligned with certification.

Why people pick it:

• Runs on existing Windows/Linux VMs or cloud images instead of a new appliance SKU
• Teams Direct Routing + SIP provider templates cut the “blank IP Group” blank-page problem
• Good fit for labs, mid-size tenants, and MSPs who want repeatable software images

What still bites: HA, monitoring, cert renewal, and Contact/FQDN discipline are still your problem. A wizard gets you Active OPTIONS faster; it does not invent dial plans or voice routing policies. Pair it with dial plan vs voice routing policy once the trunk is green.

Decision tree that survives a design review

1. Already own one? Stay on that vendor unless you have a hard support or feature gap. Migration cost beats “shiny certified other.”
2. Need appliance + field familiarity? AudioCodes Mediant is still the path of least resistance in a lot of enterprise DR shops.
3. Standardized on Ribbon elsewhere? Keep Ribbon; pick Edge vs Core from capacity and topology, not branding.
4. Software-only / cloud VM / fast template standup? anynode is usually the shortlist winner — still validate HA and ops runbooks.
5. Multi-tenant hosting? Vendor choice does not replace Microsoft’s hosted domain model. Read hosted SBC Direct Routing domain before you invent FQDNs.

Same Teams rules no matter which box you buy

Certification does not forgive bad Direct Routing hygiene. On every vendor you’ll still need:

• Public FQDN (not *.onmicrosoft.com), DNS to the SBC, and a cert whose CN/SAN matches what you send in Contact
• New-CsOnlinePSTNGateway (or derived trunks in hosted models) with the right signaling port
• Conscious media bypass / LMO choices — see Media Bypass vs Local Media Optimization
• Dial plans and voice routing policies that actually send calls to that gateway

When things break, isolate the bucket before you blame the brand:

• Inactive / TLS / cert / DNS → stay on signaling identity
• SIP 403 / wrong tenant → Contact and domain story (hosted setups especially)
• SIP 488 → SDP, codecs, SRTP — 488 Not Acceptable Here
• One-way audio / quality → media path and bypass — Direct Routing troubleshooting hub

Operator Connect is a different product path if you’re comparing “run my own SBC” vs “carrier owns the SBC.” Background: Direct Routing vs Operator Connect.

Minimal Teams pairing reminder (vendor-neutral)

New-CsOnlinePSTNGateway -FQDN sbc.contoso.com -SipSignalingPort 5061 -Enabled $true
Get-CsOnlinePSTNGateway | Select-Object Identity, Fqdn, SipSignalingPort, Enabled, MediaBypass

Port and TLS details follow the vendor note. Don’t copy a blog’s 5061/5067 blindly if your SBC signaling interface listens elsewhere. Confirm OPTIONS Active before you chase number patterns.

Checklist before you freeze the PO

1. Product + firmware appear on Microsoft’s certified Direct Routing table?
2. Media bypass / LMO required for your sites — and supported on that exact SKU?
3. Who opens the vendor ticket first when Microsoft asks for an SBC investigation report?
4. Backup, HA, and cert renewal owned by a named team?
5. SIP trunk provider has a tested interop guide for that SBC?
6. Your Teams voice design (routes, usages, dial plans) drafted separately from the SBC brand choice?